Prev: Principles for AI in the public sector
To ensure these principles are put into practice, UNISON is advocating for:
Public service staff must have a meaningful say in how AI affects their work and the services they provide. This should include:
- A legal duty on public sector employers to consult trade unions before introducing high-risk and intrusive forms of AI
- Consultation that begins at the formative stage of any proposal to introduce new technology in public service delivery
- Meaningful involvement of frontline staff in the design, procurement, trial, implementation, and maintenance of AI systems in the public sector
- Rights for public service workers to challenge decisions about service delivery made by AI systems
Strong guardrails are needed to ensure AI is implemented responsibly in our public services. We need:
- Employment legislation to protect public service workers from the potential harms of algorithmic decision-making
- Properly resourced public sector regulators with the capacity to effectively oversee AI systems
- Strict ethical, environmental and transparency conditions for companies providing AI to local authorities, the NHS, schools, and other public services
Public service workers need protection and control over how their data is used in AI systems. UNISON is asking for:
- Mandatory disclosure to public service workers about what data is being collected in their workplace
- The right for public sector staff to access their personal data and any inferences drawn from it
- Protection against excessive surveillance and monitoring in public service environments
- Joint control and access to data generated in public services
Full visibility is critical for building trust in how AI is used across public services. We need:
- Mandatory reporting of AI use across all levels of government and public services
- Public sector procurement standards that include joint access to data and full contract transparency
- Published environmental and equality impact assessments of AI services implemented in the public sector
Workers should have ongoing input into how AI systems operate in public services. This should include:
- Formal involvement of public service workers and unions at all stages of AI system design and deployment
- Regular review and assessment of systems with public service worker participation
- Clear lines of accountability for AI systems and their outputs in public service delivery
Public service workers need the skills to work confidently with AI systems. UNISON is asking for:
- Training for all public service employees who interact with AI so they can use tools effectively and critically evaluate outputs
- Education for public sector managers and decision-makers in AI principles to enable responsible procurement
- Mandatory people plans when introducing new technologies in public services, with investment in worker development
- Economy-wide programmes to support public service workers in adapting to technological change
Public sector organisations need resources and guidance to implement AI responsibly. UNISON is asking for:
- Guidance and frameworks from central government to support all levels of public services
- Collaborative knowledge-sharing between public sector organisations
- Resources for smaller public bodies to develop AI expertise
