Eighteen Governments Offer Non-Binding Recommendations for AI Security
Every now and then, global governments humbly suggest that Big Tech companies might adjust their actions ever so slightly to avoid existential harm if it’s not too much trouble. A group of eighteen countries including the U.S. and Britain took one of those tentative steps this week by offering non-binding recommendations for building AI systems that are “secure by design”. It’s not clear whether the Tech Overlords took notice.