14 October 2026

AI engineers argue agents need well-placed controls, not more restrictions

• At the AI Engineer New York conference, speakers argued that safe AI agents depend on where controls sit.
• Where controls and human judgment are placed matters more than how many restrictions are added.
• The lesson was drawn from frontier AI labs and financial firms that presented at the event.

How it was covered

Latent Spaceswyx & Alessio

Conference theme, agent autonomy versus safety, placement of controls and human judgment