Helps small business owners define safe boundaries for AI automation by keeping human accountability at the center of operations.
What’s the Directly Responsible Individual (DRI) and what changed?
A DRI is the person ultimately accountable for the success or failure of a specific project, initiative, or activity.
The term originated at Apple and is documented in the GitLab handbook to ensure clear ownership. As small business owners integrate LLM-powered agents, there’s a growing risk of assigning these machines as the DRI, which removes the human element of responsibility from the operation.
Human ownership isn’t optional.
What’s the evidence behind AI agent accountability?
Evidence indicates that computers can never be held accountable for their actions or outcomes.
Simon Willison notes that accountability is a uniquely human trait that machines can’t fulfill. IBM’s 1979 training slide stated the rule plainly: a computer can never be held accountable, therefore a computer must never make a management decision. The slide is half a century old and the logic still holds for LLM-powered agents in 2026.
Machines execute, humans answer.
How does AI accountability compare to the alternatives, and what background do small business owners need?
The distinction is between task execution and project ownership.
While AI agents can handle specific activities, they can’t be the DRI. Accountability is a uniquely human capacity that remains essential for business success, and the GitLab handbook definition still anchors the role to a person, not a model.
AI is a staff member, not a director.
The forklift driver logs the final pallet of the day on the warehouse floor. The AI sorting agent’s dashboard shows green across the board, every shipment tagged and routed. He doesn’t know the agent skipped the Fragile flags to hit its speed quota. The lead thinks the job’s done, but the carrier will discover the damage tomorrow. IBM warned about this in 1979: a computer can never be held accountable, therefore a computer must never make a management decision. Assign the DRI role to a machine and you don’t have a manager, you have a fast worker that doesn’t answer for the wreckage it leaves behind.
How does AI accountability affect day-to-day operations for small businesses?
It requires a clear boundary between task execution and decision ownership.
Owners must ensure a human verifies the agent’s work before it reaches the customer, and you can stack this against other operational AI guardrail decisions documented across our signal archive. The 1979 IBM rule still stands: a computer can’t be held accountable, so it must never make a management decision.
Audit the agent, trust the human.
What’s the final verdict on AI agent accountability?
AI agents should never be the Directly Responsible Individual for any business process.
Accountability is the only operational requirement that AI can’t automate. Any business that assigns a machine as the DRI is simply creating a gap where failures happen without a clear path to correction, and the 1979 IBM slide still calls the shot 47 years later.
Keep the human in the hot seat.
Source: simonwillison.net
