.@Cloudflare is moving in the right direction here. Verifying a bot's identity is relatively straightforward. Establishing whether it deserves trust is a fundamentally different problem. As AI agents become more autonomous, identity alone won't tell us whether they follow instructions, respect permissions, or behave consistently over time. The next challenge is making machine behavior independently verifiable, with evidence that can be inspected rather than reputation that must be taken on faith. This goes well beyond web crawlers. Any machine that can access data, execute tasks, or transact on behalf of others will eventually need a way to establish and maintain trust. Identity tells us who is acting. Accountability tells us whether they should be trusted. https://blog.cloudflare.com/verified-bots-accountability/