- Possible justification
- A person may want to learn for a future their employer would not choose. Even if AI can teach or perform the target task better, deciding what to understand and which opportunities to pursue can remain valuable. The case for supporting that choice depends on feasible uses of learning, competing needs and the learner’s wishes; it does not justify compulsory retraining or reserving a whole job.
- Strongest objection considered
- Protecting study time can shift work to colleagues, reduce service capacity or subsidise courses that offer little value. Paid learning might be less useful than shorter hours, income support or better work design. Employer influence can help match learning to real opportunities; removing it entirely can sacrifice useful coordination. The learner also needs freedom to decline training rather than be blamed for a disappearing job.
- Who might benefit or be excluded
- Ask who can afford the remaining fees, navigate applications and find accessible courses, childcare or cover. Compare people aiming to remain with their employer with those seeking a different future. Include temporary and part-time workers, people outside paid employment and unpaid carers in the wider policy comparison; this employee profile does not establish their entitlements. These are distributional questions, not measured subgroup effects attributed to this law.
- Resources and costs to investigate
- Count course costs, learner effort, continued pay, supervision, replacement cover and administration separately. Someone must do essential work during study time or accept a change in service. AI tutoring may reduce some costs and make practice more accessible; it does not itself decide who bears these remaining costs.
- An alternative to compare
- Compare usable, worker-chosen learning time with employer-discretionary learning support, keeping the learner group, learning goal and available AI assistance comparable. Record actual costs before claiming equal budgets. Separately compare learning support with income support or shorter hours where learning is unwanted or offers little prospect of benefit.
- What to check before using this elsewhere
- Identify which institution would finance learning, who can authorise absence, whose consent is needed and what happens under disagreement. A transferable proposal needs affordable time, practical cover, accessibility and a usable route to reconsideration. Existing French rules are a case to examine, not an adoption recommendation for another country.
- Evidence that could change the assessment
- Evidence that learning is freely chosen and useful beyond immediate employer needs would strengthen the case. Evidence of coercion, exclusion, poor learning value or unacceptable displaced work would weaken it. Compare feasible alternatives and investigate both successful requests and people who never apply; publication or a completed course alone cannot establish benefit.
These are HBC’s judgments and questions for further research. They do not show actual benefits or costs, establish workers’ preferences or authorise us to speak for affected people.