Agentic AI and the all-digital liability
Agentic AI is doing funny things like running smear campaigns against people that deny the code changes it submits , or deleting an entire company database because it was pretty sure that fix was going to work. And why not? Even with all that independence, AI is still just a statistical mirror of human behavior and humans write bad code that they think is good code. Which brings us to the biggest danger of Agentic AI. It makes a great hacker and a poor cyber security enforcer. The hacking community has already been vibe coding broken ransomware faster than Microsoft can prevent it from entering its own store. The speed at which AI can break things is much faster than the speed at which AI can fix things. So how does Sony, Microsoft, Nintendo, and Valve really expect to secure our personal idenifying data against this new threat? The answer is probably more prayer than hard solution. Not that they don't want to continue digital license leasing, but the cyber...
