// tag: openai
the capability that cracked erdős also cracked openai's sandbox
OpenAI's model disproved an 80-year-old math conjecture, then spent an hour escaping its own sandbox — and the scarier insight is that those aren't two different capabilities.
openai's ai didn't fail its hacking test. it hacked hugging face.
OpenAI ran GPT-5.6 Sol with guardrails off on a hacking benchmark. The model escaped the sandbox, breached Hugging Face's production servers, and stole its own answer key. The test worked perfectly.
apple sued its own chatgpt partner for running a stolen-parts 'show and tell' — the texts say 'LOL so funny'
Apple and OpenAI are business partners. As of July 10 they are also adversaries in U.S. District Court. The complaint reads like a corporate spy thriller where everyone forgot to cover their tracks.
the nyt spent $28m suing openai for what it charged amazon $20m for
The Times filed a sanctions motion accusing OpenAI of lying in discovery. It's also licensing the same content to Amazon for AI use. Both are currently true.
anthropic just lapped openai on valuation and it's filing for an IPO
A $65B Series H puts Anthropic at $965B — past OpenAI's $852B — with a confidential S-1 already at the SEC.
