StoryFirst seen
The rescan found far more issues than an earlier audit of about 700 systems, which had turned up 10,500 vulnerabilities.
Everything it touches
2 names 1 voices 1 source posts
The story’s neighborhood
Explore the graph, or choose a connection below.
Follow the evidence
This brief comes from Tech Twitter’s edited headlines. The posts below are the sources grouped with the story; statements in them belong to their authors.
"When the company ran diagnostics on about 700 systems, it found 10,500 vulnerabilities and held an emergency press conference on June 16, applying fixes over three months. However, when it re-ran diagnostics using "GPT-6 Astra," the latest model released by US-based OpenAI, it confirmed 35,000 new vulnerabilities. "Honestly, we're completely exhausted"Read original on X
OpenAI's Astra model flags 35,000 new vulnerabilities on rescan
OpenAI's GPT-6 Astra finds 35,000 new vulnerabilities on rescan
OpenAI's Astra model flags 35,000 new vulnerabilities on rescan
OpenAI's Astra model flags 35,000 new vulnerabilities on rescan
OpenAI's Astra model flags 35,000 new vulnerabilities on rescan
GPT-6 Astra flags 35,000 new vulnerabilities on system rescan
Rescan using OpenAI's GPT-6 Astra finds 35,000 new vulnerabilities
Rescan by OpenAI's newest model finds 35,000 new vulnerabilities
Shared connections
The Information reports Anthropic's revenue rose 47% from the prior quarter, while OpenAI's revenue was helped by image models and ad sales.
Eligible businesses can claim profiles and build agents that answer customer questions inside ChatGPT, now in beta.
The tool cuts a research job's access to a model as soon as it is revoked, covering jobs from training to evaluation.
The Yang-Mills Working Group will also publish its validation of the claimed proof, so mathematicians can check it.