Artificial intelligence (AI) is going from assisting humans with work to making active decisions on their behalf. That’s thanks to agentic AI, autonomous agents that can act, decide, negotiate, transact or perform tasks on behalf of human users or organizations. Agentic AI is currently in its pilot stage, but according to Gartner, 33% of enterprise applications could be incorporating it by 2028. As someone deeply involved in the evolution of Web2 and Web3, I see this transition as a fundamental change in how digital identity will work. I also see an essential piece of the puzzle still lacking attention: identity. Agentic AI’s potential doesn’t just hinge on what it can do but who it represents, how it’s verified and how it recognizes other systems and agents.

Full opinion : Agentic AI’s potential doesn’t just hinge on what it can do but who it represents.