AGI

AGI is the hypothetical point where an AI system can match or beat humans across most cognitive work, not just narrow tasks. It does not exist today; every deployed system, however impressive, is narrow. The term matters because it anchors both the hype and the safety debate, and because vendors sometimes borrow its gravity to sell ordinary automation. A grounded security program plans for the AI that exists, not the one in the headlines.