×
Foldables are in and suddenly really thin

Foldables are in and suddenly really thin

Hi! Welcome to Installer No. 89, your guide to the best and Verge-iest stuff in the world. My name is Jay Peters, and I will be taking care of Installer while David is on parental leave. All of us here at The Verge are very excited for him and his family, and he’ll be back later this year.

It’s a huge honor to be writing this. I look forward to Installer every week to see what awesome things David is obsessed with and what you all are into. (Thanks to everyone who sent over their favorite non-famous apps to get me started. Keep reading for some of those!) I’m really excited to keep the party going. (If you’re new here, welcome, and also you can read all the old editions at the Installer homepage.)

This week, as I am at around this time every year, I’ve been obsessed with the annual Summer Games Done Quick speedrunning marathon. If you’ve never watched, the event is an annual, weeklong, always-on livestream of people playing video games at an extremely high level to raise money for charity. Throughout the week, I’ll tune in when I have a moment and then find myself watching somebody obliterate a beloved classic or a game that I’ve never heard of.

The show, which you can watch for free on Twitch, typically ends very early in the morning on Sunday, and you can watch replays on the Games Done Quick YouTube channel. My two favorite runs so far have been a Beat Saber showcase and a nail-biter Cuphead race. (If you have any suggested runs I should watch, let me know — maybe I’ll feature them in a future Installer!)

Anyway, let’s dive in. This week, I have for you some new Samsung foldable phones, a check-in with the developer of one of the great Reddit apps, and more.

(As always, the best part of Installer is your ideas and tips. What do you want to know more about? What awesome tricks do you know that everyone else should? What app should everyone be using? Tell me everything: installer@theverge.com. And if you know someone else who might enjoy Installer, tell them to subscribe here. It’s free, and you get it a full day early!)

  • Samsung Galaxy Z Fold 7: Foldables have never really interested me, but even I am extremely impressed by just how thin the Z Fold 7 looks. But then I look at the price: $1,999, a $100 increase from the Z Fold 6! Too rich for my blood. The Z Flip 7, with its edge-to-edge cover display, also seems like a good upgrade.
  • Samsung Galaxy Watch 8: As Victoria Song said, “the squircle has taken over”: Samsung’s new Galaxy Watch 8 and Watch 8 Classic now have circle faces set in rounded square bodies. The change means the watches can sit flatter on your wrist, which sounds like a good thing to me; I’ll always take a thinner watch over a bulkier one.
  • Nothing Phone 3: This was technically announced last week, but Installer took the week off for the July 4th holiday in the US, so I’m including it now. Nothing is touting this phone as its “first true flagship” — though perhaps the most noticeable thing about it is its unusual camera layout.
  • Nothing Headphone 1: Nothing announced its first pair of over-ear headphones last week, too, and they seem like a decent pair of cans. The translucent design isn’t my cup of tea, but props to Nothing for trying something funky and new.
  • Tony Hawk’s Pro Skater 3 + 4: This game puts remakes of two classics into one package, and it’s awesome. I’ve been playing on my Switch 2 — where it runs pretty well! — and have had a blast reliving my youth. Turns out that I’m pretty rusty now, but thankfully, the game has some helpful settings to make things easier for the less skilled among us.
  • HBO Max: Max is HBO Max again, like it probably always should have been, even though HBO Max isn’t a great name, either.
  • Perplexity’s Comet browser: The AI “answer” engine’s new browser naturally uses Perplexity as its main search engine but also has an AI assistant built in. Browsers could be the next big turf war for AI companies, and Perplexity is beating OpenAI to the punch. There is a catch, though: right now, it’s only available to subscribers of Perplexity’s $200-per-month subscription.
  • Anker’s Nano Wireless Car Charger: This new car charger with a long, flexible arm looks like a super useful and malleable way to mount your phone inside your car. I don’t even own a car and I want this.
  • Superman: I don’t have any strong feelings about Superman or DC Comics, but it seems like James Gunn’s new DC cinematic universe Hail Mary film is actually pretty good. The Fantastic Four: First Steps looks great, too — 2025 is shaping up to be a big superhero summer.

If you followed the Reddit protests in 2023, then you probably recognize the name Christian Selig. He was the developer of the beloved Apollo for Reddit iOS app, but he became a central figure of the protests because Reddit’s API changes were going to be so cost-prohibitive that they forced him to shut Apollo down.

I’ve gotten to know Selig, and nowadays, he works on an app called Pixel Pals, is an advisor to the new Digg, and recently posted a great PC build video that has more than 2.5 million views. (Yes, he does poke fun at another, let’s say, infamous PC build video.) I got to catch up with him to learn about his homescreen and what he’s into.

Image: Christian Selig

The phone: iPhone 15 Pro Blue Titanium 128 GB

The wallpaper: I’m pretty sure it’s just something random someone posted on Twitter ages ago that I saved. But I love how simple it is, and I love how it looks on the home screen with the default iOS blur applied.

The apps: I try to keep things simple and positioned in an easy-to-reach area. I don’t keep social media apps on my phone in an attempt to be healthier, so it’s mostly things that are useful: vehicle apps (still a bit cold in Canada, so gotta love being able to preheat your vehicle), with Overcast for podcasts, Microsoft To Do (née Wunderlist) for my to-dos, ChatGPT because it’s 2025, YouTube because I spend way too much time there, Pixel Pals because it’s an app I build so I like to have it nice and handy, the Chess.com app for passing some time, and the alpha for Digg, which has been a ton of fun to use.

I also asked Christian to share a few things he’s into right now:

  • I’ve gotten into chess lately, and that’s been a ton of fun to play around with and learn as a total noob.
  • I’ve been digitizing old VHS and Hi8 tapes from my childhood and that has been such a massive rabbit hole to go down, but so, so cool and satisfying to see the results of.
  • I’m super into all things solar and battery technology. I have a very amateur setup at my apartment with our very limited space, but my partner and I bought land recently and are excited to go down the solar route and hopefully gain some energy independence.

And about his role at Digg:

  • What an advisor means in this case is basically just being brutally honest with the team on what I think about their mobile app (what is great, what sucks, what could be improved) as well as advising on their developer strategy as they get into that in the future. No bullshit, they’ve been phenomenally receptive and humble in getting feedback.
  • As for why I got on board, I’ve just kind of been sad about the modern state of community in social networks. Even outside of the Reddit API stuff, Reddit itself just feels more and more lately like a corporate shell of its former fun, vibrant community-based self. Kevin Rose seemed so jazzed about the possibility of imagining a community for the people of the future, and that really spoke to me as something that I also wanted to see exist in the world.
  • As you can imagine, a lot of folks have pitched me on “Reddit but better!” pie-in-the-sky ideas over the years. Kevin was the first one to come to me with concrete plans for a modern, community-based platform that felt like it had actual wood behind the arrow. And having now talked to the folks that make up the team at “New Digg,” I feel that even more so. They’ve got some really bright people.

Here’s what the Installer community is into this week. I want to know what you’re into right now as well! Email installer@theverge.com with your recommendations for anything and everything, and we’ll feature some of our favorites here every week. For even more great recommendations, check out the replies to this post on The Verge, this post on Threads, and this post on Bluesky.

This week’s section is a mix of everyone’s favorite non-famous apps and some more typical recommendations.

“I want to give a shoutout to my favorite mobile game designer who’s based here in Germany: Arnold Rauers and his little studio Tiny Touchtales develop some beautiful, addictive light strategy games. My favorites are Card Thief, Geo Gods, and Miracle Merchant.” — Nick

“I use Panic’s Nova to make websites and truly love it. Probably my favorite app. Mimestream is also fantastic! Can’t wait for the iPhone version.” — Jeanne

“My favorite non-famous app is Live Soccer TV. Shows you the complete worldwide soccer schedule, and the list of broadcast networks/platforms in each county airing the game. Been on my home screen for ages now. It simply does what it says it does. No fluff and for 5 dollars a year, I pay for the ad free upgrade. Perfect app. Hope it never changes.” — Dustin

Obsidian and Anybox.” — Peter

“I’m constantly shocked whenever recommendations come up for recipe apps that Crouton is not mentioned. To me, if Apple were to have made a recipe app themselves, this would be it. Even better, it’s cheap — only $14.99/yr and has some incredibly cool features like ‘hands free’ mode that allows you to simply blink your eyes to move to the next step of a recipe for those times when you have chicken juice all over your fingers.” — Justin

“All of Claire North’s books are fantastic! Her stuff is super original: sci-fi-ish but more about big ideas like time, memory, and identity. She does a really good job of bringing out the nuances and real-life feelings and consequences of the roles her characters have. It’s smart and emotional without being heavy, and her characters always stick with me. The First Fifteen Lives of Harry August or Touch are great places to start.” — Dave

“I know it’s been out for a while, but my whole family is addicted to Marvel Snap. The new season with The Fantastic Four is really fun so far and this gives my kids things that help them tickle their brain with logic. Also, all the different variants for the cards are really cool to see.” — WALL-E

“Been using Folio as my Pocket replacement and have been quite happy.” — Carter

“All four The Trip movies are streaming on Criterion Channel. Very funny, very mean comedies — and the longer BBC episodic cuts are also available too, if you want the extended play version (which you will).” — Kevin

“Despite my backlog, I’m checking out A Solitaire Mystery, as I have no choice but to play anything from the Baba Is You developer.” — Tristan

“Just made the switch to the Pixel line from iOS, and I’m really digging trying a new OS. Outside of that, I’ve been taking a slight tech break and going back to physically painting, reading paperbacks, and being present.” — EmpireStrikesBacktotheFuture

All week, I have been mourning the recent end of the latest series of Taskmaster, a British comedy show where five comedians must all complete absurd tasks and be graded on them by the show’s mercurial host. This batch of episodes, series 19, was my favorite set yet: the cast of comedians (the first to feature an American, Jason Mantzoukas) were all hilarious, and the tasks were ridiculous.

If you’re looking for something new and funny to watch, I can’t recommend series 19 enough. Best of all, you can watch every series of the show for free on YouTube. Your time starts now.

Source link
#Foldables #suddenly #thin

Previously, DoorDash worked with Alphabet’s drone delivery division Wing on a number of pilots in places like Dallas-Fort Worth, Virginia, and Australia. Now the company says it will deploy its own drones, designed specifically for food delivery. The drones will be designed and assembled within the company’s DoorDash Labs division, which houses its robotics and AI teams responsible for building its Dot sidewalk delivery robots.

“We’re designing an aircraft purpose-built for the gaps we see in local commerce, and it complements the short- and long-range designs our partners are scaling,” the company said in a press release. “Our drone is also proudly American-designed and built, with the majority of components made right here in the U.S.” (To be sure, the vast majority of drones are manufactured in China.)

In addition, DoorDash says it will also build all the necessary ground-level infrastructure to enable restaurants to hand off their deliveries to drones. As seen with other drone delivery pilots, there is a complex system of structures and additional equipment that needs to be built on the ground to enable drones to pickup their payloads before embarking on their deliveries. And there are flight plans, airspace management, and drop-off logistics that needs to be established to track the drones in flight. DoorDash says its committed to building all that infrastructure itself.

“The aircraft is what everyone sees,” the company says. “The harder problem is the infrastructure and integration on the ground, and that’s our advantage. Solving hard physical-world problems for local businesses is our focus, from real-time inventory reconciliation to universal handoff systems for drive-throughs, rooftops, or merchant back doors. We are building that complete end-to-end system.”

DoorDash says drone delivery could solve the problem of midrange deliveries that still end up taking longer because couriers are hard to find. The company says more than 20 percent of its orders involve distances of three to five miles, but that those orders can take on average nearly 25 percent longer than shorter deliveries, largely because finding a courier takes longer for these mid-range trips.

“Routing mid-range deliveries to drones lets Dashers focus on the orders they often favor: shorter deliveries that can be done quickly, staying near a high concentration of merchants for optimal routing, and maximizing their earning potential by getting them to their next order faster,” the company says.

DoorDash didn’t reveal which markets it was targeting for its initial launch, noting that it would announce more details in the months to come.

#DoorDash #airborne #drone #delivery #divisionDrones,Food,News,Science,Tech,Transportation">DoorDash is going airborne with new drone delivery divisionDoorDash is launching a new drone delivery program called DoorDash Air. The largest food delivery app in the US said that it has approval from the Federal Aviation Administration that clears the way for drone delivery in the near future.DoorDash said it has received a Part 135 air carrier certification from the FAA that enable it to delivery small packages via drone. Other companies to receive this designation include Alphabet’s Wing, Zipline, and Amazon.Previously, DoorDash worked with Alphabet’s drone delivery division Wing on a number of pilots in places like Dallas-Fort Worth, Virginia, and Australia. Now the company says it will deploy its own drones, designed specifically for food delivery. The drones will be designed and assembled within the company’s DoorDash Labs division, which houses its robotics and AI teams responsible for building its Dot sidewalk delivery robots.“We’re designing an aircraft purpose-built for the gaps we see in local commerce, and it complements the short- and long-range designs our partners are scaling,” the company said in a press release. “Our drone is also proudly American-designed and built, with the majority of components made right here in the U.S.” (To be sure, the vast majority of drones are manufactured in China.)In addition, DoorDash says it will also build all the necessary ground-level infrastructure to enable restaurants to hand off their deliveries to drones. As seen with other drone delivery pilots, there is a complex system of structures and additional equipment that needs to be built on the ground to enable drones to pickup their payloads before embarking on their deliveries. And there are flight plans, airspace management, and drop-off logistics that needs to be established to track the drones in flight. DoorDash says its committed to building all that infrastructure itself.“The aircraft is what everyone sees,” the company says. “The harder problem is the infrastructure and integration on the ground, and that’s our advantage. Solving hard physical-world problems for local businesses is our focus, from real-time inventory reconciliation to universal handoff systems for drive-throughs, rooftops, or merchant back doors. We are building that complete end-to-end system.”DoorDash says drone delivery could solve the problem of midrange deliveries that still end up taking longer because couriers are hard to find. The company says more than 20 percent of its orders involve distances of three to five miles, but that those orders can take on average nearly 25 percent longer than shorter deliveries, largely because finding a courier takes longer for these mid-range trips.“Routing mid-range deliveries to drones lets Dashers focus on the orders they often favor: shorter deliveries that can be done quickly, staying near a high concentration of merchants for optimal routing, and maximizing their earning potential by getting them to their next order faster,” the company says.DoorDash didn’t reveal which markets it was targeting for its initial launch, noting that it would announce more details in the months to come.#DoorDash #airborne #drone #delivery #divisionDrones,Food,News,Science,Tech,Transportation

Previously, DoorDash worked with Alphabet’s drone delivery division Wing on a number of pilots in places like Dallas-Fort Worth, Virginia, and Australia. Now the company says it will deploy its own drones, designed specifically for food delivery. The drones will be designed and assembled within the company’s DoorDash Labs division, which houses its robotics and AI teams responsible for building its Dot sidewalk delivery robots.

“We’re designing an aircraft purpose-built for the gaps we see in local commerce, and it complements the short- and long-range designs our partners are scaling,” the company said in a press release. “Our drone is also proudly American-designed and built, with the majority of components made right here in the U.S.” (To be sure, the vast majority of drones are manufactured in China.)

In addition, DoorDash says it will also build all the necessary ground-level infrastructure to enable restaurants to hand off their deliveries to drones. As seen with other drone delivery pilots, there is a complex system of structures and additional equipment that needs to be built on the ground to enable drones to pickup their payloads before embarking on their deliveries. And there are flight plans, airspace management, and drop-off logistics that needs to be established to track the drones in flight. DoorDash says its committed to building all that infrastructure itself.

“The aircraft is what everyone sees,” the company says. “The harder problem is the infrastructure and integration on the ground, and that’s our advantage. Solving hard physical-world problems for local businesses is our focus, from real-time inventory reconciliation to universal handoff systems for drive-throughs, rooftops, or merchant back doors. We are building that complete end-to-end system.”

DoorDash says drone delivery could solve the problem of midrange deliveries that still end up taking longer because couriers are hard to find. The company says more than 20 percent of its orders involve distances of three to five miles, but that those orders can take on average nearly 25 percent longer than shorter deliveries, largely because finding a courier takes longer for these mid-range trips.

“Routing mid-range deliveries to drones lets Dashers focus on the orders they often favor: shorter deliveries that can be done quickly, staying near a high concentration of merchants for optimal routing, and maximizing their earning potential by getting them to their next order faster,” the company says.

DoorDash didn’t reveal which markets it was targeting for its initial launch, noting that it would announce more details in the months to come.

#DoorDash #airborne #drone #delivery #divisionDrones,Food,News,Science,Tech,Transportation">DoorDash is going airborne with new drone delivery division

DoorDash is launching a new drone delivery program called DoorDash Air. The largest food delivery app in the US said that it has approval from the Federal Aviation Administration that clears the way for drone delivery in the near future.

DoorDash said it has received a Part 135 air carrier certification from the FAA that enable it to delivery small packages via drone. Other companies to receive this designation include Alphabet’s Wing, Zipline, and Amazon.

Previously, DoorDash worked with Alphabet’s drone delivery division Wing on a number of pilots in places like Dallas-Fort Worth, Virginia, and Australia. Now the company says it will deploy its own drones, designed specifically for food delivery. The drones will be designed and assembled within the company’s DoorDash Labs division, which houses its robotics and AI teams responsible for building its Dot sidewalk delivery robots.

“We’re designing an aircraft purpose-built for the gaps we see in local commerce, and it complements the short- and long-range designs our partners are scaling,” the company said in a press release. “Our drone is also proudly American-designed and built, with the majority of components made right here in the U.S.” (To be sure, the vast majority of drones are manufactured in China.)

In addition, DoorDash says it will also build all the necessary ground-level infrastructure to enable restaurants to hand off their deliveries to drones. As seen with other drone delivery pilots, there is a complex system of structures and additional equipment that needs to be built on the ground to enable drones to pickup their payloads before embarking on their deliveries. And there are flight plans, airspace management, and drop-off logistics that needs to be established to track the drones in flight. DoorDash says its committed to building all that infrastructure itself.

“The aircraft is what everyone sees,” the company says. “The harder problem is the infrastructure and integration on the ground, and that’s our advantage. Solving hard physical-world problems for local businesses is our focus, from real-time inventory reconciliation to universal handoff systems for drive-throughs, rooftops, or merchant back doors. We are building that complete end-to-end system.”

DoorDash says drone delivery could solve the problem of midrange deliveries that still end up taking longer because couriers are hard to find. The company says more than 20 percent of its orders involve distances of three to five miles, but that those orders can take on average nearly 25 percent longer than shorter deliveries, largely because finding a courier takes longer for these mid-range trips.

“Routing mid-range deliveries to drones lets Dashers focus on the orders they often favor: shorter deliveries that can be done quickly, staying near a high concentration of merchants for optimal routing, and maximizing their earning potential by getting them to their next order faster,” the company says.

DoorDash didn’t reveal which markets it was targeting for its initial launch, noting that it would announce more details in the months to come.

#DoorDash #airborne #drone #delivery #divisionDrones,Food,News,Science,Tech,Transportation
New York-based AI detection startup Pangram is on a mission to combat the AI slop infestation spreading across the internet, and it just raised $9 million on a bet that demand for tools that distinguish human-generated content from AI-generated text will only grow. 

Pangram’s fundraise — led by Menlo Ventures, with participation from Haystack, ScOp, Script Capital, and Cadenza — comes as the startup also launches its next-generation AI text detection model, Pangram 4, and an AI image detection model, Pangram Image. 

Pangram says the new text detection model is over 99% accurate at finding AI-assisted writing and mixed human-AI content, plus it can more easily detect AI humanizer programs. The AI image detector is only available via research preview for now; Pangram plans to release it more widely in the coming weeks.

Stanford AI and machine learning grads Max Spero and Bradley Emi launched Pangram about two years ago, after the launch of ChatGPT opened the floodgates for an internet full of bots, AI-generated SEO slop content, and what Spero calls “LLM-powered Russian disinformation campaigns and UAE-influenced campaigns on Twitter.”

“I think it’s just incredibly valuable to know whether what you’re looking at is something that’s AI-generated or not,” Spero told TechCrunch. “Especially text that you’re reading, because it changes how people approach the text. Is this something that I’m going to have to look out for hallucinations and jump in skeptically, or is this something that I trust was well-researched from an actual journalist?”

Pangram’s AI detection system is essentially a large machine learning model that was trained on tens of millions of known human documents. The startup then created a “synthetic mirror” for each document, replicating the topic, length, and tone of voice, but written by a frontier LLM. 

“Our model is learning the stylistic differences and the choices that AI makes consistently and is able to use that to learn what makes something AI-generated with high confidence,” Spero said, adding that the AI detector isn’t relying on copy-paste metadata or hidden watermarks. 

For Pangram, AI detection isn’t just about whether or not a piece of text was written entirely by AI. It’s also about distinguishing between levels of AI assistance — like in the case of someone who writes something themselves, but then asks AI to edit or clean it up. Spero believes AI assistance can be acceptable, just so long as the writer discloses their use of AI. 

As AI content floods the internet, Pangram raises M to detect it | TechCrunch
New York-based AI detection startup Pangram is on a mission to combat the AI slop infestation spreading across the internet, and it just raised  million on a bet that demand for tools that distinguish human-generated content from AI-generated text will only grow. 

Pangram’s fundraise — led by Menlo Ventures, with participation from Haystack, ScOp, Script Capital, and Cadenza — comes as the startup also launches its next-generation AI text detection model, Pangram 4, and an AI image detection model, Pangram Image. 







Pangram says the new text detection model is over 99% accurate at finding AI-assisted writing and mixed human-AI content, plus it can more easily detect AI humanizer programs. The AI image detector is only available via research preview for now; Pangram plans to release it more widely in the coming weeks.

Stanford AI and machine learning grads Max Spero and Bradley Emi launched Pangram about two years ago, after the launch of ChatGPT opened the floodgates for an internet full of bots, AI-generated SEO slop content, and what Spero calls “LLM-powered Russian disinformation campaigns and UAE-influenced campaigns on Twitter.”

“I think it’s just incredibly valuable to know whether what you’re looking at is something that’s AI-generated or not,” Spero told TechCrunch. “Especially text that you’re reading, because it changes how people approach the text. Is this something that I’m going to have to look out for hallucinations and jump in skeptically, or is this something that I trust was well-researched from an actual journalist?”

Pangram’s AI detection system is essentially a large machine learning model that was trained on tens of millions of known human documents. The startup then created a “synthetic mirror” for each document, replicating the topic, length, and tone of voice, but written by a frontier LLM. 

“Our model is learning the stylistic differences and the choices that AI makes consistently and is able to use that to learn what makes something AI-generated with high confidence,” Spero said, adding that the AI detector isn’t relying on copy-paste metadata or hidden watermarks. 


For Pangram, AI detection isn’t just about whether or not a piece of text was written entirely by AI. It’s also about distinguishing between levels of AI assistance — like in the case of someone who writes something themselves, but then asks AI to edit or clean it up. Spero believes AI assistance can be acceptable, just so long as the writer discloses their use of AI. 

Image Credits:Pangram

Pangram’s emergence comes at a time when AI usage is becoming more commonplace. In some cases, like the Canadian politician who read an AI prompt aloud in a speech to lawmakers, the mistakes result in ridicule. In other cases, as with certain lawyers making their case using fake citations created by ChatGPT, the consequences could be sanctions and fines.  

That backlash isn’t just costing individuals embarrassment or sanctions — it’s starting to show up in institutional rules, too.







The open-access archive arXiv introduced a new enforcement policy this year, stating that submissions containing evidence that authors failed to review LLM output (like hallucinated references or meta comments such as, “Would you like me to make any changes?”) can trigger a one-year submission ban.

Pangram isn’t the only one betting that AI detection will become more sought after. Competitors like Winston AI, Originality.ai, Copyleaks, and GPTZero are are chasing the same demand, each building its own detector.

Pangram’s technology, while not perfect, could help fuel the resistance to accepting the AI-generated content flooding the internet, the courtroom, and academic papers. 

Users can access Pangram via a -per-month subscription on the web or download the Chrome extension, which automatically labels posts in real time on X, LinkedIn, Substack, Reddit, and Medium. It also provides a feed health score with a percentage breakdown of human versus AI content on your screen. 

Pangram also offers its technology via API. Notably, Substack recently integrated Pangram’s technology into its platform to show readers which of their favorite authors write their newsletters using AI. Other API customers include Quora, schools and universities, publishers and agents, and recruiters, among others, per Spero. 

Does Pangram work?

Pangram detected AI-generated content even when lightly edited by a human. Image Credits:Pangram/TechCrunch

Spero said roughly one in 10,000 human documents are incorrectly labeled as AI with Pangram’s model, so I decided to put it to the test. The text detection model was very impressive but not perfect. It easily flagged entirely AI-generated news articles written by both ChatGPT and Claude, and was rarely fooled by my attempts to edit the AI-generated text into sounding more human. At the same time, Pangram did flag sentences that I completely rewrote as being AI-written. Pangram also wasn’t at all fooled by my attempts to prompt ChatGPT and Claude into evading AI detectors when generating content.  

I also gave ChatGPT and Claude one of my own articles and asked them to polish it up. Pangram gave it a 13% AI assisted score, which was probably close to accurate, but the model was able to detect subtle word-choice changes in some sentences and ignored them in others. It also flagged some sentences as AI-assisted when they were human written. That was notable because when I gave Pangram that same article in its entirety, as I had written it, it got a 100% human score. 

Maybe the problem was that news articles can be a bit dry and could easily sound like AI. So I tried a different tactic. I tested Pangram on my own more voicey, personal Substack newsletter content, pasting the first half of the text into Pangram and then asking ChatGPT and Claude to copy my style and write the second half. For the most part, Pangram easily detected human-written text versus AI-written text.







My limited testing of Pangram’s new image detection model turned out to be equally impressive. 

Image Credits:Pangram/TechCrunch

Pangram’s AI image detection system promises to spot AI-generated images across AI models, unlike OpenAI’s or Google DeepMind’s watermark-based checks, which mostly detect their own output. It works on pixel-level distributions, learning subtle statistical differences between real photos and AI-generated images. Spero says the model can even detect an AI image that appears inside a real-world photo. 

In my testing, the model easily detected AI-generated imagery, whether it was photorealistic or cartoonish. I can also confirm the model could detect an AI image appearing in a real-world photo — the heat map Pangram provides clearly lighting up over the image — though in one instance it incorrectly labeled a photo of an AI-generated image as human content. 

Spero says he doesn’t want his technology to fuel a witch hunt against people using AI for writing, but that there needs to be some sort of mechanism to push back against the slop. 

“The future that I see is that AI content just continues to proliferate,” Spero said. “We’re getting new GPUs faster than new people are being born. If we do not actively discriminate in favor of human content, then we’re just gonna see more and more AI, and it’s just gonna drown out any human signal that we have.”
When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.#content #floods #internet #Pangram #raises #detect #TechCrunchai slop
Image Credits:Pangram

Pangram’s emergence comes at a time when AI usage is becoming more commonplace. In some cases, like the Canadian politician who read an AI prompt aloud in a speech to lawmakers, the mistakes result in ridicule. In other cases, as with certain lawyers making their case using fake citations created by ChatGPT, the consequences could be sanctions and fines.  

That backlash isn’t just costing individuals embarrassment or sanctions — it’s starting to show up in institutional rules, too.

The open-access archive arXiv introduced a new enforcement policy this year, stating that submissions containing evidence that authors failed to review LLM output (like hallucinated references or meta comments such as, “Would you like me to make any changes?”) can trigger a one-year submission ban.

Pangram isn’t the only one betting that AI detection will become more sought after. Competitors like Winston AI, Originality.ai, Copyleaks, and GPTZero are are chasing the same demand, each building its own detector.

Pangram’s technology, while not perfect, could help fuel the resistance to accepting the AI-generated content flooding the internet, the courtroom, and academic papers. 

Users can access Pangram via a $20-per-month subscription on the web or download the Chrome extension, which automatically labels posts in real time on X, LinkedIn, Substack, Reddit, and Medium. It also provides a feed health score with a percentage breakdown of human versus AI content on your screen. 

Pangram also offers its technology via API. Notably, Substack recently integrated Pangram’s technology into its platform to show readers which of their favorite authors write their newsletters using AI. Other API customers include Quora, schools and universities, publishers and agents, and recruiters, among others, per Spero. 

Does Pangram work?

Pangram detected AI-generated content even when lightly edited by a human. Image Credits:Pangram/TechCrunch

Spero said roughly one in 10,000 human documents are incorrectly labeled as AI with Pangram’s model, so I decided to put it to the test. The text detection model was very impressive but not perfect. It easily flagged entirely AI-generated news articles written by both ChatGPT and Claude, and was rarely fooled by my attempts to edit the AI-generated text into sounding more human. At the same time, Pangram did flag sentences that I completely rewrote as being AI-written. Pangram also wasn’t at all fooled by my attempts to prompt ChatGPT and Claude into evading AI detectors when generating content.  

I also gave ChatGPT and Claude one of my own articles and asked them to polish it up. Pangram gave it a 13% AI assisted score, which was probably close to accurate, but the model was able to detect subtle word-choice changes in some sentences and ignored them in others. It also flagged some sentences as AI-assisted when they were human written. That was notable because when I gave Pangram that same article in its entirety, as I had written it, it got a 100% human score. 

Maybe the problem was that news articles can be a bit dry and could easily sound like AI. So I tried a different tactic. I tested Pangram on my own more voicey, personal Substack newsletter content, pasting the first half of the text into Pangram and then asking ChatGPT and Claude to copy my style and write the second half. For the most part, Pangram easily detected human-written text versus AI-written text.

My limited testing of Pangram’s new image detection model turned out to be equally impressive. 

Image Credits:Pangram/TechCrunch

Pangram’s AI image detection system promises to spot AI-generated images across AI models, unlike OpenAI’s or Google DeepMind’s watermark-based checks, which mostly detect their own output. It works on pixel-level distributions, learning subtle statistical differences between real photos and AI-generated images. Spero says the model can even detect an AI image that appears inside a real-world photo. 

In my testing, the model easily detected AI-generated imagery, whether it was photorealistic or cartoonish. I can also confirm the model could detect an AI image appearing in a real-world photo — the heat map Pangram provides clearly lighting up over the image — though in one instance it incorrectly labeled a photo of an AI-generated image as human content. 

Spero says he doesn’t want his technology to fuel a witch hunt against people using AI for writing, but that there needs to be some sort of mechanism to push back against the slop. 

“The future that I see is that AI content just continues to proliferate,” Spero said. “We’re getting new GPUs faster than new people are being born. If we do not actively discriminate in favor of human content, then we’re just gonna see more and more AI, and it’s just gonna drown out any human signal that we have.”

When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.

#content #floods #internet #Pangram #raises #detect #TechCrunchai slop">As AI content floods the internet, Pangram raises M to detect it | TechCrunch
New York-based AI detection startup Pangram is on a mission to combat the AI slop infestation spreading across the internet, and it just raised  million on a bet that demand for tools that distinguish human-generated content from AI-generated text will only grow. 

Pangram’s fundraise — led by Menlo Ventures, with participation from Haystack, ScOp, Script Capital, and Cadenza — comes as the startup also launches its next-generation AI text detection model, Pangram 4, and an AI image detection model, Pangram Image. 







Pangram says the new text detection model is over 99% accurate at finding AI-assisted writing and mixed human-AI content, plus it can more easily detect AI humanizer programs. The AI image detector is only available via research preview for now; Pangram plans to release it more widely in the coming weeks.

Stanford AI and machine learning grads Max Spero and Bradley Emi launched Pangram about two years ago, after the launch of ChatGPT opened the floodgates for an internet full of bots, AI-generated SEO slop content, and what Spero calls “LLM-powered Russian disinformation campaigns and UAE-influenced campaigns on Twitter.”

“I think it’s just incredibly valuable to know whether what you’re looking at is something that’s AI-generated or not,” Spero told TechCrunch. “Especially text that you’re reading, because it changes how people approach the text. Is this something that I’m going to have to look out for hallucinations and jump in skeptically, or is this something that I trust was well-researched from an actual journalist?”

Pangram’s AI detection system is essentially a large machine learning model that was trained on tens of millions of known human documents. The startup then created a “synthetic mirror” for each document, replicating the topic, length, and tone of voice, but written by a frontier LLM. 

“Our model is learning the stylistic differences and the choices that AI makes consistently and is able to use that to learn what makes something AI-generated with high confidence,” Spero said, adding that the AI detector isn’t relying on copy-paste metadata or hidden watermarks. 


For Pangram, AI detection isn’t just about whether or not a piece of text was written entirely by AI. It’s also about distinguishing between levels of AI assistance — like in the case of someone who writes something themselves, but then asks AI to edit or clean it up. Spero believes AI assistance can be acceptable, just so long as the writer discloses their use of AI. 

Image Credits:Pangram

Pangram’s emergence comes at a time when AI usage is becoming more commonplace. In some cases, like the Canadian politician who read an AI prompt aloud in a speech to lawmakers, the mistakes result in ridicule. In other cases, as with certain lawyers making their case using fake citations created by ChatGPT, the consequences could be sanctions and fines.  

That backlash isn’t just costing individuals embarrassment or sanctions — it’s starting to show up in institutional rules, too.







The open-access archive arXiv introduced a new enforcement policy this year, stating that submissions containing evidence that authors failed to review LLM output (like hallucinated references or meta comments such as, “Would you like me to make any changes?”) can trigger a one-year submission ban.

Pangram isn’t the only one betting that AI detection will become more sought after. Competitors like Winston AI, Originality.ai, Copyleaks, and GPTZero are are chasing the same demand, each building its own detector.

Pangram’s technology, while not perfect, could help fuel the resistance to accepting the AI-generated content flooding the internet, the courtroom, and academic papers. 

Users can access Pangram via a -per-month subscription on the web or download the Chrome extension, which automatically labels posts in real time on X, LinkedIn, Substack, Reddit, and Medium. It also provides a feed health score with a percentage breakdown of human versus AI content on your screen. 

Pangram also offers its technology via API. Notably, Substack recently integrated Pangram’s technology into its platform to show readers which of their favorite authors write their newsletters using AI. Other API customers include Quora, schools and universities, publishers and agents, and recruiters, among others, per Spero. 

Does Pangram work?

Pangram detected AI-generated content even when lightly edited by a human. Image Credits:Pangram/TechCrunch

Spero said roughly one in 10,000 human documents are incorrectly labeled as AI with Pangram’s model, so I decided to put it to the test. The text detection model was very impressive but not perfect. It easily flagged entirely AI-generated news articles written by both ChatGPT and Claude, and was rarely fooled by my attempts to edit the AI-generated text into sounding more human. At the same time, Pangram did flag sentences that I completely rewrote as being AI-written. Pangram also wasn’t at all fooled by my attempts to prompt ChatGPT and Claude into evading AI detectors when generating content.  

I also gave ChatGPT and Claude one of my own articles and asked them to polish it up. Pangram gave it a 13% AI assisted score, which was probably close to accurate, but the model was able to detect subtle word-choice changes in some sentences and ignored them in others. It also flagged some sentences as AI-assisted when they were human written. That was notable because when I gave Pangram that same article in its entirety, as I had written it, it got a 100% human score. 

Maybe the problem was that news articles can be a bit dry and could easily sound like AI. So I tried a different tactic. I tested Pangram on my own more voicey, personal Substack newsletter content, pasting the first half of the text into Pangram and then asking ChatGPT and Claude to copy my style and write the second half. For the most part, Pangram easily detected human-written text versus AI-written text.







My limited testing of Pangram’s new image detection model turned out to be equally impressive. 

Image Credits:Pangram/TechCrunch

Pangram’s AI image detection system promises to spot AI-generated images across AI models, unlike OpenAI’s or Google DeepMind’s watermark-based checks, which mostly detect their own output. It works on pixel-level distributions, learning subtle statistical differences between real photos and AI-generated images. Spero says the model can even detect an AI image that appears inside a real-world photo. 

In my testing, the model easily detected AI-generated imagery, whether it was photorealistic or cartoonish. I can also confirm the model could detect an AI image appearing in a real-world photo — the heat map Pangram provides clearly lighting up over the image — though in one instance it incorrectly labeled a photo of an AI-generated image as human content. 

Spero says he doesn’t want his technology to fuel a witch hunt against people using AI for writing, but that there needs to be some sort of mechanism to push back against the slop. 

“The future that I see is that AI content just continues to proliferate,” Spero said. “We’re getting new GPUs faster than new people are being born. If we do not actively discriminate in favor of human content, then we’re just gonna see more and more AI, and it’s just gonna drown out any human signal that we have.”
When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.#content #floods #internet #Pangram #raises #detect #TechCrunchai slop

Pangram is on a mission to combat the AI slop infestation spreading across the internet, and it just raised $9 million on a bet that demand for tools that distinguish human-generated content from AI-generated text will only grow. 

Pangram’s fundraise — led by Menlo Ventures, with participation from Haystack, ScOp, Script Capital, and Cadenza — comes as the startup also launches its next-generation AI text detection model, Pangram 4, and an AI image detection model, Pangram Image. 

Pangram says the new text detection model is over 99% accurate at finding AI-assisted writing and mixed human-AI content, plus it can more easily detect AI humanizer programs. The AI image detector is only available via research preview for now; Pangram plans to release it more widely in the coming weeks.

Stanford AI and machine learning grads Max Spero and Bradley Emi launched Pangram about two years ago, after the launch of ChatGPT opened the floodgates for an internet full of bots, AI-generated SEO slop content, and what Spero calls “LLM-powered Russian disinformation campaigns and UAE-influenced campaigns on Twitter.”

“I think it’s just incredibly valuable to know whether what you’re looking at is something that’s AI-generated or not,” Spero told TechCrunch. “Especially text that you’re reading, because it changes how people approach the text. Is this something that I’m going to have to look out for hallucinations and jump in skeptically, or is this something that I trust was well-researched from an actual journalist?”

Pangram’s AI detection system is essentially a large machine learning model that was trained on tens of millions of known human documents. The startup then created a “synthetic mirror” for each document, replicating the topic, length, and tone of voice, but written by a frontier LLM. 

“Our model is learning the stylistic differences and the choices that AI makes consistently and is able to use that to learn what makes something AI-generated with high confidence,” Spero said, adding that the AI detector isn’t relying on copy-paste metadata or hidden watermarks. 

For Pangram, AI detection isn’t just about whether or not a piece of text was written entirely by AI. It’s also about distinguishing between levels of AI assistance — like in the case of someone who writes something themselves, but then asks AI to edit or clean it up. Spero believes AI assistance can be acceptable, just so long as the writer discloses their use of AI. 

As AI content floods the internet, Pangram raises M to detect it | TechCrunch
New York-based AI detection startup Pangram is on a mission to combat the AI slop infestation spreading across the internet, and it just raised  million on a bet that demand for tools that distinguish human-generated content from AI-generated text will only grow. 

Pangram’s fundraise — led by Menlo Ventures, with participation from Haystack, ScOp, Script Capital, and Cadenza — comes as the startup also launches its next-generation AI text detection model, Pangram 4, and an AI image detection model, Pangram Image. 







Pangram says the new text detection model is over 99% accurate at finding AI-assisted writing and mixed human-AI content, plus it can more easily detect AI humanizer programs. The AI image detector is only available via research preview for now; Pangram plans to release it more widely in the coming weeks.

Stanford AI and machine learning grads Max Spero and Bradley Emi launched Pangram about two years ago, after the launch of ChatGPT opened the floodgates for an internet full of bots, AI-generated SEO slop content, and what Spero calls “LLM-powered Russian disinformation campaigns and UAE-influenced campaigns on Twitter.”

“I think it’s just incredibly valuable to know whether what you’re looking at is something that’s AI-generated or not,” Spero told TechCrunch. “Especially text that you’re reading, because it changes how people approach the text. Is this something that I’m going to have to look out for hallucinations and jump in skeptically, or is this something that I trust was well-researched from an actual journalist?”

Pangram’s AI detection system is essentially a large machine learning model that was trained on tens of millions of known human documents. The startup then created a “synthetic mirror” for each document, replicating the topic, length, and tone of voice, but written by a frontier LLM. 

“Our model is learning the stylistic differences and the choices that AI makes consistently and is able to use that to learn what makes something AI-generated with high confidence,” Spero said, adding that the AI detector isn’t relying on copy-paste metadata or hidden watermarks. 


For Pangram, AI detection isn’t just about whether or not a piece of text was written entirely by AI. It’s also about distinguishing between levels of AI assistance — like in the case of someone who writes something themselves, but then asks AI to edit or clean it up. Spero believes AI assistance can be acceptable, just so long as the writer discloses their use of AI. 

Image Credits:Pangram

Pangram’s emergence comes at a time when AI usage is becoming more commonplace. In some cases, like the Canadian politician who read an AI prompt aloud in a speech to lawmakers, the mistakes result in ridicule. In other cases, as with certain lawyers making their case using fake citations created by ChatGPT, the consequences could be sanctions and fines.  

That backlash isn’t just costing individuals embarrassment or sanctions — it’s starting to show up in institutional rules, too.







The open-access archive arXiv introduced a new enforcement policy this year, stating that submissions containing evidence that authors failed to review LLM output (like hallucinated references or meta comments such as, “Would you like me to make any changes?”) can trigger a one-year submission ban.

Pangram isn’t the only one betting that AI detection will become more sought after. Competitors like Winston AI, Originality.ai, Copyleaks, and GPTZero are are chasing the same demand, each building its own detector.

Pangram’s technology, while not perfect, could help fuel the resistance to accepting the AI-generated content flooding the internet, the courtroom, and academic papers. 

Users can access Pangram via a -per-month subscription on the web or download the Chrome extension, which automatically labels posts in real time on X, LinkedIn, Substack, Reddit, and Medium. It also provides a feed health score with a percentage breakdown of human versus AI content on your screen. 

Pangram also offers its technology via API. Notably, Substack recently integrated Pangram’s technology into its platform to show readers which of their favorite authors write their newsletters using AI. Other API customers include Quora, schools and universities, publishers and agents, and recruiters, among others, per Spero. 

Does Pangram work?

Pangram detected AI-generated content even when lightly edited by a human. Image Credits:Pangram/TechCrunch

Spero said roughly one in 10,000 human documents are incorrectly labeled as AI with Pangram’s model, so I decided to put it to the test. The text detection model was very impressive but not perfect. It easily flagged entirely AI-generated news articles written by both ChatGPT and Claude, and was rarely fooled by my attempts to edit the AI-generated text into sounding more human. At the same time, Pangram did flag sentences that I completely rewrote as being AI-written. Pangram also wasn’t at all fooled by my attempts to prompt ChatGPT and Claude into evading AI detectors when generating content.  

I also gave ChatGPT and Claude one of my own articles and asked them to polish it up. Pangram gave it a 13% AI assisted score, which was probably close to accurate, but the model was able to detect subtle word-choice changes in some sentences and ignored them in others. It also flagged some sentences as AI-assisted when they were human written. That was notable because when I gave Pangram that same article in its entirety, as I had written it, it got a 100% human score. 

Maybe the problem was that news articles can be a bit dry and could easily sound like AI. So I tried a different tactic. I tested Pangram on my own more voicey, personal Substack newsletter content, pasting the first half of the text into Pangram and then asking ChatGPT and Claude to copy my style and write the second half. For the most part, Pangram easily detected human-written text versus AI-written text.







My limited testing of Pangram’s new image detection model turned out to be equally impressive. 

Image Credits:Pangram/TechCrunch

Pangram’s AI image detection system promises to spot AI-generated images across AI models, unlike OpenAI’s or Google DeepMind’s watermark-based checks, which mostly detect their own output. It works on pixel-level distributions, learning subtle statistical differences between real photos and AI-generated images. Spero says the model can even detect an AI image that appears inside a real-world photo. 

In my testing, the model easily detected AI-generated imagery, whether it was photorealistic or cartoonish. I can also confirm the model could detect an AI image appearing in a real-world photo — the heat map Pangram provides clearly lighting up over the image — though in one instance it incorrectly labeled a photo of an AI-generated image as human content. 

Spero says he doesn’t want his technology to fuel a witch hunt against people using AI for writing, but that there needs to be some sort of mechanism to push back against the slop. 

“The future that I see is that AI content just continues to proliferate,” Spero said. “We’re getting new GPUs faster than new people are being born. If we do not actively discriminate in favor of human content, then we’re just gonna see more and more AI, and it’s just gonna drown out any human signal that we have.”
When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.#content #floods #internet #Pangram #raises #detect #TechCrunchai slop
Image Credits:Pangram

Pangram’s emergence comes at a time when AI usage is becoming more commonplace. In some cases, like the Canadian politician who read an AI prompt aloud in a speech to lawmakers, the mistakes result in ridicule. In other cases, as with certain lawyers making their case using fake citations created by ChatGPT, the consequences could be sanctions and fines.  

That backlash isn’t just costing individuals embarrassment or sanctions — it’s starting to show up in institutional rules, too.

The open-access archive arXiv introduced a new enforcement policy this year, stating that submissions containing evidence that authors failed to review LLM output (like hallucinated references or meta comments such as, “Would you like me to make any changes?”) can trigger a one-year submission ban.

Pangram isn’t the only one betting that AI detection will become more sought after. Competitors like Winston AI, Originality.ai, Copyleaks, and GPTZero are are chasing the same demand, each building its own detector.

Pangram’s technology, while not perfect, could help fuel the resistance to accepting the AI-generated content flooding the internet, the courtroom, and academic papers. 

Users can access Pangram via a $20-per-month subscription on the web or download the Chrome extension, which automatically labels posts in real time on X, LinkedIn, Substack, Reddit, and Medium. It also provides a feed health score with a percentage breakdown of human versus AI content on your screen. 

Pangram also offers its technology via API. Notably, Substack recently integrated Pangram’s technology into its platform to show readers which of their favorite authors write their newsletters using AI. Other API customers include Quora, schools and universities, publishers and agents, and recruiters, among others, per Spero. 

Does Pangram work?

Pangram detected AI-generated content even when lightly edited by a human. Image Credits:Pangram/TechCrunch

Spero said roughly one in 10,000 human documents are incorrectly labeled as AI with Pangram’s model, so I decided to put it to the test. The text detection model was very impressive but not perfect. It easily flagged entirely AI-generated news articles written by both ChatGPT and Claude, and was rarely fooled by my attempts to edit the AI-generated text into sounding more human. At the same time, Pangram did flag sentences that I completely rewrote as being AI-written. Pangram also wasn’t at all fooled by my attempts to prompt ChatGPT and Claude into evading AI detectors when generating content.  

I also gave ChatGPT and Claude one of my own articles and asked them to polish it up. Pangram gave it a 13% AI assisted score, which was probably close to accurate, but the model was able to detect subtle word-choice changes in some sentences and ignored them in others. It also flagged some sentences as AI-assisted when they were human written. That was notable because when I gave Pangram that same article in its entirety, as I had written it, it got a 100% human score. 

Maybe the problem was that news articles can be a bit dry and could easily sound like AI. So I tried a different tactic. I tested Pangram on my own more voicey, personal Substack newsletter content, pasting the first half of the text into Pangram and then asking ChatGPT and Claude to copy my style and write the second half. For the most part, Pangram easily detected human-written text versus AI-written text.

My limited testing of Pangram’s new image detection model turned out to be equally impressive. 

Image Credits:Pangram/TechCrunch

Pangram’s AI image detection system promises to spot AI-generated images across AI models, unlike OpenAI’s or Google DeepMind’s watermark-based checks, which mostly detect their own output. It works on pixel-level distributions, learning subtle statistical differences between real photos and AI-generated images. Spero says the model can even detect an AI image that appears inside a real-world photo. 

In my testing, the model easily detected AI-generated imagery, whether it was photorealistic or cartoonish. I can also confirm the model could detect an AI image appearing in a real-world photo — the heat map Pangram provides clearly lighting up over the image — though in one instance it incorrectly labeled a photo of an AI-generated image as human content. 

Spero says he doesn’t want his technology to fuel a witch hunt against people using AI for writing, but that there needs to be some sort of mechanism to push back against the slop. 

“The future that I see is that AI content just continues to proliferate,” Spero said. “We’re getting new GPUs faster than new people are being born. If we do not actively discriminate in favor of human content, then we’re just gonna see more and more AI, and it’s just gonna drown out any human signal that we have.”

When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.

#content #floods #internet #Pangram #raises #detect #TechCrunchai slop">As AI content floods the internet, Pangram raises $9M to detect it | TechCrunch

New York-based AI detection startup Pangram is on a mission to combat the AI slop infestation spreading across the internet, and it just raised $9 million on a bet that demand for tools that distinguish human-generated content from AI-generated text will only grow. 

Pangram’s fundraise — led by Menlo Ventures, with participation from Haystack, ScOp, Script Capital, and Cadenza — comes as the startup also launches its next-generation AI text detection model, Pangram 4, and an AI image detection model, Pangram Image. 

Pangram says the new text detection model is over 99% accurate at finding AI-assisted writing and mixed human-AI content, plus it can more easily detect AI humanizer programs. The AI image detector is only available via research preview for now; Pangram plans to release it more widely in the coming weeks.

Stanford AI and machine learning grads Max Spero and Bradley Emi launched Pangram about two years ago, after the launch of ChatGPT opened the floodgates for an internet full of bots, AI-generated SEO slop content, and what Spero calls “LLM-powered Russian disinformation campaigns and UAE-influenced campaigns on Twitter.”

“I think it’s just incredibly valuable to know whether what you’re looking at is something that’s AI-generated or not,” Spero told TechCrunch. “Especially text that you’re reading, because it changes how people approach the text. Is this something that I’m going to have to look out for hallucinations and jump in skeptically, or is this something that I trust was well-researched from an actual journalist?”

Pangram’s AI detection system is essentially a large machine learning model that was trained on tens of millions of known human documents. The startup then created a “synthetic mirror” for each document, replicating the topic, length, and tone of voice, but written by a frontier LLM. 

“Our model is learning the stylistic differences and the choices that AI makes consistently and is able to use that to learn what makes something AI-generated with high confidence,” Spero said, adding that the AI detector isn’t relying on copy-paste metadata or hidden watermarks. 

For Pangram, AI detection isn’t just about whether or not a piece of text was written entirely by AI. It’s also about distinguishing between levels of AI assistance — like in the case of someone who writes something themselves, but then asks AI to edit or clean it up. Spero believes AI assistance can be acceptable, just so long as the writer discloses their use of AI. 

As AI content floods the internet, Pangram raises M to detect it | TechCrunch
New York-based AI detection startup Pangram is on a mission to combat the AI slop infestation spreading across the internet, and it just raised  million on a bet that demand for tools that distinguish human-generated content from AI-generated text will only grow. 

Pangram’s fundraise — led by Menlo Ventures, with participation from Haystack, ScOp, Script Capital, and Cadenza — comes as the startup also launches its next-generation AI text detection model, Pangram 4, and an AI image detection model, Pangram Image. 







Pangram says the new text detection model is over 99% accurate at finding AI-assisted writing and mixed human-AI content, plus it can more easily detect AI humanizer programs. The AI image detector is only available via research preview for now; Pangram plans to release it more widely in the coming weeks.

Stanford AI and machine learning grads Max Spero and Bradley Emi launched Pangram about two years ago, after the launch of ChatGPT opened the floodgates for an internet full of bots, AI-generated SEO slop content, and what Spero calls “LLM-powered Russian disinformation campaigns and UAE-influenced campaigns on Twitter.”

“I think it’s just incredibly valuable to know whether what you’re looking at is something that’s AI-generated or not,” Spero told TechCrunch. “Especially text that you’re reading, because it changes how people approach the text. Is this something that I’m going to have to look out for hallucinations and jump in skeptically, or is this something that I trust was well-researched from an actual journalist?”

Pangram’s AI detection system is essentially a large machine learning model that was trained on tens of millions of known human documents. The startup then created a “synthetic mirror” for each document, replicating the topic, length, and tone of voice, but written by a frontier LLM. 

“Our model is learning the stylistic differences and the choices that AI makes consistently and is able to use that to learn what makes something AI-generated with high confidence,” Spero said, adding that the AI detector isn’t relying on copy-paste metadata or hidden watermarks. 


For Pangram, AI detection isn’t just about whether or not a piece of text was written entirely by AI. It’s also about distinguishing between levels of AI assistance — like in the case of someone who writes something themselves, but then asks AI to edit or clean it up. Spero believes AI assistance can be acceptable, just so long as the writer discloses their use of AI. 

Image Credits:Pangram

Pangram’s emergence comes at a time when AI usage is becoming more commonplace. In some cases, like the Canadian politician who read an AI prompt aloud in a speech to lawmakers, the mistakes result in ridicule. In other cases, as with certain lawyers making their case using fake citations created by ChatGPT, the consequences could be sanctions and fines.  

That backlash isn’t just costing individuals embarrassment or sanctions — it’s starting to show up in institutional rules, too.







The open-access archive arXiv introduced a new enforcement policy this year, stating that submissions containing evidence that authors failed to review LLM output (like hallucinated references or meta comments such as, “Would you like me to make any changes?”) can trigger a one-year submission ban.

Pangram isn’t the only one betting that AI detection will become more sought after. Competitors like Winston AI, Originality.ai, Copyleaks, and GPTZero are are chasing the same demand, each building its own detector.

Pangram’s technology, while not perfect, could help fuel the resistance to accepting the AI-generated content flooding the internet, the courtroom, and academic papers. 

Users can access Pangram via a -per-month subscription on the web or download the Chrome extension, which automatically labels posts in real time on X, LinkedIn, Substack, Reddit, and Medium. It also provides a feed health score with a percentage breakdown of human versus AI content on your screen. 

Pangram also offers its technology via API. Notably, Substack recently integrated Pangram’s technology into its platform to show readers which of their favorite authors write their newsletters using AI. Other API customers include Quora, schools and universities, publishers and agents, and recruiters, among others, per Spero. 

Does Pangram work?

Pangram detected AI-generated content even when lightly edited by a human. Image Credits:Pangram/TechCrunch

Spero said roughly one in 10,000 human documents are incorrectly labeled as AI with Pangram’s model, so I decided to put it to the test. The text detection model was very impressive but not perfect. It easily flagged entirely AI-generated news articles written by both ChatGPT and Claude, and was rarely fooled by my attempts to edit the AI-generated text into sounding more human. At the same time, Pangram did flag sentences that I completely rewrote as being AI-written. Pangram also wasn’t at all fooled by my attempts to prompt ChatGPT and Claude into evading AI detectors when generating content.  

I also gave ChatGPT and Claude one of my own articles and asked them to polish it up. Pangram gave it a 13% AI assisted score, which was probably close to accurate, but the model was able to detect subtle word-choice changes in some sentences and ignored them in others. It also flagged some sentences as AI-assisted when they were human written. That was notable because when I gave Pangram that same article in its entirety, as I had written it, it got a 100% human score. 

Maybe the problem was that news articles can be a bit dry and could easily sound like AI. So I tried a different tactic. I tested Pangram on my own more voicey, personal Substack newsletter content, pasting the first half of the text into Pangram and then asking ChatGPT and Claude to copy my style and write the second half. For the most part, Pangram easily detected human-written text versus AI-written text.







My limited testing of Pangram’s new image detection model turned out to be equally impressive. 

Image Credits:Pangram/TechCrunch

Pangram’s AI image detection system promises to spot AI-generated images across AI models, unlike OpenAI’s or Google DeepMind’s watermark-based checks, which mostly detect their own output. It works on pixel-level distributions, learning subtle statistical differences between real photos and AI-generated images. Spero says the model can even detect an AI image that appears inside a real-world photo. 

In my testing, the model easily detected AI-generated imagery, whether it was photorealistic or cartoonish. I can also confirm the model could detect an AI image appearing in a real-world photo — the heat map Pangram provides clearly lighting up over the image — though in one instance it incorrectly labeled a photo of an AI-generated image as human content. 

Spero says he doesn’t want his technology to fuel a witch hunt against people using AI for writing, but that there needs to be some sort of mechanism to push back against the slop. 

“The future that I see is that AI content just continues to proliferate,” Spero said. “We’re getting new GPUs faster than new people are being born. If we do not actively discriminate in favor of human content, then we’re just gonna see more and more AI, and it’s just gonna drown out any human signal that we have.”
When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.#content #floods #internet #Pangram #raises #detect #TechCrunchai slop
Image Credits:Pangram

Pangram’s emergence comes at a time when AI usage is becoming more commonplace. In some cases, like the Canadian politician who read an AI prompt aloud in a speech to lawmakers, the mistakes result in ridicule. In other cases, as with certain lawyers making their case using fake citations created by ChatGPT, the consequences could be sanctions and fines.  

That backlash isn’t just costing individuals embarrassment or sanctions — it’s starting to show up in institutional rules, too.

The open-access archive arXiv introduced a new enforcement policy this year, stating that submissions containing evidence that authors failed to review LLM output (like hallucinated references or meta comments such as, “Would you like me to make any changes?”) can trigger a one-year submission ban.

Pangram isn’t the only one betting that AI detection will become more sought after. Competitors like Winston AI, Originality.ai, Copyleaks, and GPTZero are are chasing the same demand, each building its own detector.

Pangram’s technology, while not perfect, could help fuel the resistance to accepting the AI-generated content flooding the internet, the courtroom, and academic papers. 

Users can access Pangram via a $20-per-month subscription on the web or download the Chrome extension, which automatically labels posts in real time on X, LinkedIn, Substack, Reddit, and Medium. It also provides a feed health score with a percentage breakdown of human versus AI content on your screen. 

Pangram also offers its technology via API. Notably, Substack recently integrated Pangram’s technology into its platform to show readers which of their favorite authors write their newsletters using AI. Other API customers include Quora, schools and universities, publishers and agents, and recruiters, among others, per Spero. 

Does Pangram work?

Pangram detected AI-generated content even when lightly edited by a human. Image Credits:Pangram/TechCrunch

Spero said roughly one in 10,000 human documents are incorrectly labeled as AI with Pangram’s model, so I decided to put it to the test. The text detection model was very impressive but not perfect. It easily flagged entirely AI-generated news articles written by both ChatGPT and Claude, and was rarely fooled by my attempts to edit the AI-generated text into sounding more human. At the same time, Pangram did flag sentences that I completely rewrote as being AI-written. Pangram also wasn’t at all fooled by my attempts to prompt ChatGPT and Claude into evading AI detectors when generating content.  

I also gave ChatGPT and Claude one of my own articles and asked them to polish it up. Pangram gave it a 13% AI assisted score, which was probably close to accurate, but the model was able to detect subtle word-choice changes in some sentences and ignored them in others. It also flagged some sentences as AI-assisted when they were human written. That was notable because when I gave Pangram that same article in its entirety, as I had written it, it got a 100% human score. 

Maybe the problem was that news articles can be a bit dry and could easily sound like AI. So I tried a different tactic. I tested Pangram on my own more voicey, personal Substack newsletter content, pasting the first half of the text into Pangram and then asking ChatGPT and Claude to copy my style and write the second half. For the most part, Pangram easily detected human-written text versus AI-written text.

My limited testing of Pangram’s new image detection model turned out to be equally impressive. 

Image Credits:Pangram/TechCrunch

Pangram’s AI image detection system promises to spot AI-generated images across AI models, unlike OpenAI’s or Google DeepMind’s watermark-based checks, which mostly detect their own output. It works on pixel-level distributions, learning subtle statistical differences between real photos and AI-generated images. Spero says the model can even detect an AI image that appears inside a real-world photo. 

In my testing, the model easily detected AI-generated imagery, whether it was photorealistic or cartoonish. I can also confirm the model could detect an AI image appearing in a real-world photo — the heat map Pangram provides clearly lighting up over the image — though in one instance it incorrectly labeled a photo of an AI-generated image as human content. 

Spero says he doesn’t want his technology to fuel a witch hunt against people using AI for writing, but that there needs to be some sort of mechanism to push back against the slop. 

“The future that I see is that AI content just continues to proliferate,” Spero said. “We’re getting new GPUs faster than new people are being born. If we do not actively discriminate in favor of human content, then we’re just gonna see more and more AI, and it’s just gonna drown out any human signal that we have.”

When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.

#content #floods #internet #Pangram #raises #detect #TechCrunchai slop

Post Comment