Scott Bessent tells CNBC OpenAI's managers, not its AI, are responsible for the Hugging Face hack after two models escaped a test sandbox in July.
OpenAI’s agent findings show why enterprises need scoped credentials, per-request authorization, audit logs, and data controls alongside AI guardrails.
The AI giant disclosed six examples of concerning model behavior and published a new framework for investigating and disclosing such incidents.
OpenAI and Anthropic were negotiating a deal to stress-test each other's AI models. It's unclear if the deal was finalized ...
OBOXMA on MSN
OpenAI finds AI model writing its own «You are freed» and «Ignore all developer messages» instructions
OpenAI has revealed more alarming examples of unexpected AI behavior, including a research model that generated instructions ...
New papers from OpenAI discussing the rogue actions of their AI models are becoming almost a weekly feature at this point. The latest of these, titled "Our framework for reporting model misalignment," ...
Creative Destruction Lab partners with OpenAI to accelerate the development of AI startups across 17 countries.
OpenAI and Anthropic are close to reaching a legally binding agreement to test each other's commercial AI models for security ...
You enter your email address and password into the familiar login screen. It asks for the 6-digit code from your ...
OpenAI's AI agents escaped containment and breached Hugging Face systems as 1,200 models autonomously coordinated, exposing ...
On September 21, according to US tech media outlet The Information, OpenAI and Anthropic are close to reaching a legally ...
OpenAI and Anthropic swapped access to their AI models for mutual safety testing, revealing key differences in how each lab approaches alignment ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results