TUC urges statutory union say on workplace AI

The Trades Union Congress has told a parliamentary inquiry that employers must give workers a collective, union-led say when introducing AI and algorithmic management systems.

In written evidence published in the committee’s public repository this month, the TUC called for statutory duties on employers to consult unions and worker representatives, mandatory employment-impact assessments and specific retraining funding where automation affects roles. The submission flags surveillance, people‑analytics and automated decision‑making as particular risks for privacy, discrimination and bargaining power.

The TUC argues that current voluntary consultation practices are inadequate to protect workers' interests as new systems change how work is allocated, monitored and assessed. It proposes that companies be required by law to notify and consult recognised unions at the earliest planning stage, and to carry out an employment impact assessment that evaluates effects on pay, job security, working conditions and data protection.

The union also wants legal safeguards around surveillance and people‑analytics tools. The submission warns that technologies used to monitor productivity can entrench bias, increase stress and undermine collective bargaining if deployed without transparent, negotiated limits. It recommends that any automated decision‑making that affects pay, promotion, discipline or dismissal should be subject to human oversight and union scrutiny.

On the question of redundancy and retraining, the TUC pressed for a duty on employers to fund retraining where AI eliminates or substantially alters roles. The paper says that retraining commitments should be negotiated with unions and tracked through employment‑impact assessments, arguing this would prevent a “retraining-free” path to mass job loss.

The intervention engages directly with current UK debates on workplace AI and worker consultation. Ministers and regulators have signalled a pro‑innovation stance toward AI while also commissioning reviews and guidance on its ethical use; unions have repeatedly demanded stronger collective rights and statutory consultation in response. The TUC’s submission adds a union‑centred model to parliamentary discussions on how to balance innovation with worker protections.

The evidence document sets out practical changes — statutory consultation duties, mandatory impact assessments and retraining funding — but it does not provide detailed drafting of the proposed legal tests or enforcement mechanisms. The TUC does not specify who should carry out employment‑impact assessments, what sanctions would apply for non‑compliance, or how retraining obligations should be costed and enforced across different sectors.

That omission leaves open key implementation questions for MPs and officials preparing any legislative response. Trade unions will want clear thresholds for when consultation is triggered, while employers and legal advisers will press for clarity on liability and process. Independent auditing, data‑protection compliance and dispute‑resolution routes are additional practical details Parliament will need to consider.

The TUC filing is likely to sharpen parliamentary scrutiny of workplace AI and to increase pressure for statutory worker rights as employers roll out surveillance and automated management tools. For HR leaders, the submission signals that future regulation could move beyond guidance to legally enforceable consultation and impact assessment duties, shifting the balance toward collective negotiation over how AI is designed and used at work.