Wiz says Project Atlas beats frontier cyber models with an agent system
The reported 90.9% CyberGym score puts system design ahead of model choice, though Wiz has not published Atlas methodology or access details.
By Ryan Merket · Published
Primary source: Forbes
Why it matters
Project Atlas supports a broader shift in AI security: orchestration, validation and proprietary context are becoming the durable assets, while the underlying models remain replaceable inputs.
On July 27th, Assaf Rappaport's Wiz introduced Project Atlas, a vulnerability research system that assigns different parts of a security investigation to specialized AI agents and models, according to Forbes. Wiz told Forbes that Atlas scored 90.9% on CyberGym, ahead of published results from Anthropic, OpenAI and Microsoft.…
When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.