Defending the internet from slop Tag "@gptzeroai" in 🧵 to detect AI on X
Joined April 2023
- Tweets6.2K
- Following50
- Followers2.5K
- Likes663
GPTZero retweeted
We just ran every United Nations General Assembly speech through GPTZero.
Roughly 1 in 5 words were AI-generated.
1 in 4 speeches were AI-generated.
1 in 3 English-language speeches were AI-generated.
GPTZero retweeted
Foreign leaders are using identical sentences in their UN speeches, almost down to the word. Coincidence or AI?
Here are three examples that @GPTZeroAI found 1/
politico.com/newsletters/wes…
GPTZero retweeted
After months of heads-down grinding, I’m proud to share that the @GPTZeroAI team has released the world’s most accurate AI detection model, GPTZero 4o.
It’s called 4o because there are four zeros: we mark less than one in 10000 human texts as AI, evaluated across diverse passages of student writing. At the same time, we detect Claude Fable with 98.4% accuracy.
We outperform Pangram in 3rd party benchmarks. Along the way, our research found major biases in the Pangram detector.
GPTZero retweeted
Análisis de discursos impartidos en UNGA encuentra múltiples instancias de posible uso de IA en su escritura, según @GPTZeroAI.
Por ej, Tanzania, Canadá, Irlanda y Noruega usaron la misma fórmula discursiva sobre el aumento en los niveles del mar.
gptzero.me/news/presidents-s…
GPTZero retweeted
KPMG report contained AI hallucinations on benefits of . . . AI ft.trib.al/2wyH5lr
GPTZero retweeted
Today, we’re launching AI Reviewer Expert Feedback.
We partnered with Emmy-winning writers to train AIs to give expert writing feedback through GPTZero.
GPTZero retweeted
Lot's of people are lauding @pangram as a tool for AI detection.
However, in the limited testing I've done I've found it shockingly easy to make minor changes to AI output and trick it into thinking it's 100% human generated.
GPTZero retweeted
Hey Alex, huge fans, and kudos for an awesome study on AI detection reliability in the ecosystem
HUGE CLARIFICATION
we were benchmarked in this study without accounting for our three-output structure, excluding our third 'mixed' class
it's like having a two party election, forgetting there's a smaller third independent party; we care about the third option because more writing on the internet is now humanized, modified, lightly edited with ai, or mixed and combined.
accounting for this, our Recall is 99.3% and FPR is 0.5%, outperforming Pangram here.
GPTZero retweeted
Check out how @GPTZeroAI now helps students sharpen their writing skills while giving teachers smarter ways to teach.
The @observer has more here 👇
observer.com/2025/08/gptzero…
#ResponsibleAI #uncorkcap
GPTZero retweeted
NCHE confirms it doesn’t recognize AIU degrees. Multiple sources in 2024 report that Prof Bazeyo’s PhD from AIU is unrecognized and widely challenged, sparking legal and internal controversies at Makerere. [kmaupdates.com/2024/03/25/pr…]
More sources: app.gptzero.me/documents/af9…
GPTZero retweeted
The thread claims Reuters was paid millions for “large scale social deception.” Recent 2025 sources (RT, AA, News-Pravda) report DoD contracts for projects labeled ASED/LSD, fueling debate via Musk and Trump statements, though independent verification remains sparse. rt.com/news/612663-musk-reut… news-pravda.com/world/2025/0…
More sources: app.gptzero.me/documents/2e4…
GPTZero retweeted
tech can also do something good. we're launching the first fact check bot on X/twitter today.
tag @gptzeroAI fact check on any thread and try it out 🙏
GPTZero retweeted
Exclusive: A @GPTZeroAI analysis shows that AI writing is commonplace in top Substack newsletters: wired.com/story/substacks-wr…
GPTZero retweeted
This is why we work with data labelers to filter out AI-generated content from crowd workers at @GPTZeroAI
Our customer’s LLMs output less hallucinations and better alignment when their data has been filtered from AI, great study from @NicolasPapernot
The internet is becoming awash in AI-made words and image. And they're increasingly likely to be used by AI in the future, leading to worse results. Here’s a look at what happens when a system is trained on its own output, over and over again.
nyti.ms/471kCip
GPTZero retweeted
Garbage In, Garbage Out: Perplexity Spreads Misinformation From Spammy AI Blog Posts trib.al/xhypT6s
GPTZero retweeted
Stoked to announce our new 10M Series A on TC, and to be working together with the amazing @nbt @jameskimEDU and @reachfund in defining the future of AI adoption.
Same day, our team is releasing a new product — GPTZero docs, the future of transparent writing with built in AI detection and writing replays. (try @ gptzero.me)
GPTZero’s founders, still in their 20s, have a profitable AI detection startup, millions in the bank and a new $10M Series A tcrn.ch/45qkCaT
GPTZero retweeted
Searches related to AI news often returned AI generated sources, reflecting more of the internet being AI generated.
Even one of Perplexity's own recommended searches on Google's Eric Schmidt returned a source that was clearly AI generated.
Link to the search: perplexity.ai/search/Eric-Sc…
Link to the source: app.gptzero.me/app/documents…
