Get the latest tech news

Search-capable AI agents may cheat on benchmark tests


None

Get the Android app

Or read this on The Register

Read more on:

Photo of Search

Search

Photo of benchmark tests

benchmark tests

Photo of capable AI agents

capable AI agents

Related news:

News photo

Call off the search - Verizon has THE Google Pixel 10 Pro deal to beat

News photo

AI Mode in Search gets new agentic features and expands globally

News photo

Show HN: Luminal – Open-source, search-based GPU compiler