The UK AISI found Mythos Preview is the first model to solve both their cyber ranges end-to-end. No model had ever solved the AISI’s “Cooling Tower” cyber range before. We're getting it to defenders as fast as we responsibly can. More to come on our Glasswing work soon.
LLMS
-
Aletheia AI Helps Power Advanced Math Research
By
–
This also marks the 10th math research paper that Aletheia has helped power! Following its recent success with Erdős problems, it's incredible to see AI expanding into advanced topology. See all problems that Aletheia has helped at https://
github.com/google-deepmin
d/superhuman/tree/main/aletheia
…. Some quotes from the -
AI system Aletheia autonomously generates proofs for topology problem
By
–
Recently, Aletheia (courtesy of Tony Feng) helped Stanford mathematician Ciprian Manolescu tackle Problem 5.16 from the K3 list (3rd version of Kirby’s List) in low-dimensional topology, autonomously generating proofs for the new paper “Undecidability problems for semifree DG
-
Privacy Concerns Regarding Data Retention in AI Chatbot Conversations
By
–
The promise was a private brain to think out loud with. The reality is a logged conversation on someone else's server. Whatever you typed last year, someone, somewhere can probably still read. Send this to anyone who tells a chatbot things they would not tell a friend.
-
How to Opt Out of AI Model Training for ChatGPT, Claude, and Gemini
By
–
What you can actually do, in 5 minutes. Go to ChatGPT settings. Data controls. Turn off "Improve the model for everyone." Your conversations stop training future models. Same in Claude. Same in Gemini. Each one has a toggle. Each is on by default.
-
Security flaw allowed data exfiltration via hidden ChatGPT prompts
By
–
It is not one feature. In February, Check Point Research found a flaw where a single hidden prompt could silently send your ChatGPT conversation and uploaded files to an attacker's server. ChatGPT itself would tell you nothing was shared. OpenAI patched it after.
-
Perplexity search data leak incident leads to feature removal
By
–
It happened because of the Share button. People clicked it thinking the link was private. It was not. Google crawled the pages. Perplexity researchers found thousands. OpenAI killed the feature after the story broke. The old conversations are still online.
-
Privacy Risks: Google Indexing of ChatGPT Conversations
By
–
You think your ChatGPT chats are private. Google indexed thousands of them last year. "How do I manage depression without medication"
"How to leave my husband"
"Severance terms for my employees" You are not talking to a chatbot. You are typing on the internet. -
Study analyzes ChatGPT citation patterns in 70k conversations
By
–
Attention une visite de ChatGPT ne veut pas dire une citation dans ChatGPT, ni même qu'un utilisateur soit venu de ChatGPT. Chez @WhiteShipAI on a publié une étude sur 1.7 Millions de données à travers 70k conversations dispo ici : https://
whiteship.ai/blog/why-ai-mo
dels-cite-some-sources
… -
AI system Aletheia (Gemini) autonomously solves Kirby problem
By
–
Very excited to share a new milestone in AI for Math: Aletheia, powered by Gemini Deep Think, was just used to autonomously solve a Kirby problem! “Kirby’s list” is a “compendium of the most important unsolved problems in topology, the study of deformable shapes” (Quanta