AWS proposes Trust in Depth framework for agent autonomy
An AWS Machine Learning Blog post by Shawn Bice, vice president of AI Services, introduces a framework called Trust in Depth that would let AI agents act on their own when their actions can be verified, reserving human judgment for decisions that need it. The post frames trust as four lifecycle questions — scoping the agent before it runs, knowing its identity and authorization at deployment, bounding what it says and does while running, and reconstructing what happened afterward — and maps them to AWS Agent Registry, AgentCore agent identity and policy enforcement, Amazon Bedrock Guardrails and automated reasoning checks. It argues that content filtering, testing or monitoring alone cannot promise an agent will behave correctly next time, so limits should be enforced deterministically outside the agent's reasoning to hold even when it is buggy, manipulated or hit by prompt injection.
AWS is packaging agent identity, enforceable policy and formally checkable rules into a single governance stance, which could shape the trust bar enterprises set before granting agents more autonomy.
In every conversation I have with enterprise leaders, they say the same thing. They want agents to do more, but they don’t trust them to do it right all the time.
That’s fair. The same creativity and autonomy that makes agents useful makes them hard to trust. Give an agent a task and it finds its own path to finish it, picking tools and sequencing steps no developer scripted. Run the same task twice and it might take a different route each time. A test that passes today tells you little about what the agent does tomorrow.
The usual answer is a person in the loop, approving everything the agent does. That works at small scale, but as agents take on more work, human approvals pile up and review fatigue sets in. People asked to check everything stop catching the things that matter. Caution becomes a bottleneck, and agents capable of far more sit idle on the valuable work that could be making more of an impact.
The fix is to let the agent act on its own when its actions can be verified, and reserve human judgment for the decisions that need it. As agents take on higher stakes work, like financial transactions, a wrong action becomes increasingly harmful. Organizations need the confidence to hand an AI agent important work, knowing it will stay within the limits set, even when no one is watching. That confidence must be built.
#AWS#Amazon Bedrock AgentCore#AWS Agent Registry#Amazon Bedrock Guardrails#Model Context Protocol