10 · Keeping a human accountable

As you automate more, it gets dangerously easy to let the system run and stop paying attention — until something goes wrong and everyone points at the tool. The final operations discipline ties the whole course together: a human is always accountable for what your automation does. A machine can act, but it can't be answerable. You can.

What "keeping a human accountable" actually means:

• Name the owner. For every important automation, one person is responsible for it working right and for fixing it when it doesn't. "It's automated" is not an owner.
• Keep a human on the decisions that matter. Money, customer promises, anything sensitive or hard to undo — a person approves or decides, the automation just assists (Lessons 4 and 8). NIST's framework is built on this: trustworthy AI is accountable, with human oversight designed in (NIST, 2023).
• Check the work. Review what your automations are doing on a regular schedule, not just when a customer complains. Catch the drift early.
• Don't hide behind the bot. If your automation makes a mistake — a wrong charge, a bad reply, a leaked detail — you own it, you fix it, and you make it right with the customer. The FTC has been clear that "the AI did it" is not a defense for misleading customers (FTC, 2023).

Why this protects your business, not just your customers:

• It keeps your brand trustworthy — people forgive an honest small business that owns its mistakes; they don't forgive being deceived or ignored.
• It keeps you legal — you, not the tool, are responsible for truthful claims and for protecting data.
• It keeps you in control — you're running the automation, not the other way around.

The whole course in one line: automate the boring, repetitive work so you have more time for what matters — but keep a human mapping it, checking it, and answerable for it. AI assists; you decide.

Check yourself. What does it mean to "name the owner" of an automation, and why is "the AI did it" never a valid excuse to a customer?

Sources

• National Institute of Standards and Technology. (2023). AI Risk Management Framework (AI RMF 1.0) — accountability and human oversight as traits of trustworthy AI. https://www.nist.gov/itl/ai-risk-management-framework
• Federal Trade Commission. (2023). Keep your AI claims in check — businesses remain responsible for what their AI does; "the AI did it" is not a defense. https://www.ftc.gov/business-guidance/blog/2023/02/keep-your-ai-claims-check