Companies & Workplaces

What you can ask today

Whether you are an employee, a manager, or in HR leadership, these are the questions worth raising:

• Does AI training at your company cover all employees, or only engineers and technical teams? The knowledge gap is not confined to technical roles.
• Are people being trained to use AI effectively in their specific workflows, or is training a generic one-time session that does not change how anyone actually works?
• Does training include how to evaluate AI outputs critically: recognising when AI is wrong, when it is reproducing bias, and when it should not be trusted?
• What is the company's plan for employees whose roles are changing because of AI? Are they being trained to work with it, or managed out on the assumption that AI will cover the gap?
• What is your company's AI training standard? Who designed it, and who verified it was delivered?

What you can do this week

• Ask HR for the company's AI training policy, in writing. Ask specifically: who does it cover, what standard does it meet, and who verified delivery? A policy that cannot be produced in written form, or that covers only engineering teams, is a gap worth naming. The response tells you more than the policy itself.
• In your next team meeting, spend 15 minutes on visibility. Ask each person to name one AI tool they use regularly and what they use it for. No formal agenda, no evaluation; just shared awareness. It makes the knowledge gap inside the team visible to the people in it.
• If your company has cited AI in any recent restructuring or redundancies, ask what training was offered to affected employees beforehand. In most jurisdictions this is not yet a legal requirement; in the EU it will become one under the AI Act. Asking the question now, and noting the answer, matters for both workers and the organisation.
• Propose a single structured hour with your direct team on AI use in your specific workflows. Make it specific: each person walks through one task they currently do manually that AI could support. That focus is where training produces real productivity change.

What you can push for

• AI training for all employees, not just engineers. A marketing coordinator who knows how to use AI tools will produce more and better work. An HR manager who understands algorithmic hiring will make better decisions. An accountant who can evaluate AI-generated analysis is less likely to sign off on something wrong. The productivity case is not confined to technical roles.
• Training that covers practical use and critical evaluation together. An employee who uses AI without understanding how it can fail is a liability. An employee who understands its limits produces reliable output. These are the same training goal.
• A formal AI training policy with defined scope, a responsible owner, and a review cycle. A one-page statement or a link to an online course is a different thing.
• Transparency about AI in employment decisions. Workers should know when AI systems are used to evaluate them, on what basis, and how to contest a decision.
• Support for legislation requiring employer investment in worker AI training, especially in sectors with high displacement risk. Colorado, Illinois, and New York have models worth building on.

Know your rights

In the EU, the AI Act classifies employment AI as high-risk. Employers must conduct risk assessments and workers have transparency rights. In the US, Colorado, Illinois, and New York have enacted protections against AI-driven discrimination in employment, including rights to notification and appeal. Most US workers currently have no equivalent protection.

If you are subject to an AI-driven employment decision you believe is wrong or discriminatory, document it. In jurisdictions with protections, you may have the right to request a human review.

Go deeper

• EU AI Act: Article 4 (AI literacy obligation for staff), Article 50 (transparency), and the employment provisions. See research file.
• SHRM: AI governance in employment as a top 2026 issue: useful framing for HR and compliance audiences. See research file.
• WEF: Eight myths sabotaging AI governance: addresses the assumption that technical teams alone can manage AI risk across an organisation. See research file.
• US state AI laws taking effect in 2026: Colorado, Illinois, Texas, California worker protections.