AI agents can't be trusted, so don't give them dangerous powers

Source: The Register — read the full report at the original publisher.

This is a curated wire item. The Continuum Brief does not republish full third-party articles; this entry links to the original source.