What to Do When Your Employer Introduces AI

Start with purpose and scope

Request a practical explanation of the intended outcome: faster drafting, lower error rates, new customer capacity or reduced cost. Ask which tasks are in scope now and which decisions remain with people.

A clear scope protects both the worker and the project. It allows training, testing and workload to be discussed against a defined change rather than a broad instruction to use AI.

Questions an experienced worker should ask

Use calm, operational questions. They reveal whether governance exists without assuming either that the tool is harmless or that jobs have already been decided.

• Which information may enter the system, and which tool or account is approved?
• Who checks output, signs decisions and handles complaints or incidents?
• How will quality, rework, workload and time saved be measured?
• Will prompts, activity, output or performance data be used to monitor individuals?
• What training, adjustment period and route for raising concerns will be provided?

Use a pilot to make hidden work visible

Bring real exceptions and downstream consequences into testing. Track the checking effort and the work created by correcting, documenting or explaining an output.

Suggest a joint review point where the team can decide which tasks to assist, automate or leave unchanged. NIST's framework supports a cycle of governing, mapping, measuring and managing rather than one approval at launch.

Know when to use a formal route

Concerns about personal data, monitoring, discrimination, accessibility, professional duties or contract changes may engage specific national law or workplace procedure. The applicable rights differ by country, sector and employment status.

Keep dates, written instructions, policies and examples. Speak to a union representative, employee forum, data-protection contact, professional body or qualified adviser where appropriate. This article is general information, not legal advice.