AI Literacy and Competency for Safety, Health, Risk, and Ergonomics Professionals

Determine whether a task, tool, data set, and professional authority are appropriate for AI use before information is entered into a system.

Characterize an AI-enabled system, including its model, application, data, function, autonomy, permissions, integrations, operating context, limitations, and foreseeable misuse.

Identify how AI may introduce or modify safety, health, ergonomic, psychosocial, organizational, privacy, security, and equity risks.

Evaluate data quality, labels, performance metrics, uncertainty, subgroup results, failure modes, bias, security risks, and local validation evidence.

Conduct an AI Use-Case Impact and Control Assessment covering value, non-AI alternatives, affected workers, evidence, human factors, controls, residual risk, and decision authority.

Design a bounded, source-grounded, and secure AI-assisted work protocol with structured outputs, abstention rules, test cases, verification steps, and qualified human review.

Verify AI-generated claims, calculations, references, recommendations, and system-performance evidence before relying on them in professional work.

Establish effective human oversight, worker consultation, contestability, documentation, escalation, and professional-boundary controls.

Develop a pilot and monitoring plan with acceptance criteria, accountable roles, fallback arrangements, management-of-change triggers, stop-use conditions, and incident-response actions.

Communicate and defend an AI-related recommendation while clearly stating the evidence, uncertainty, limitations, required controls, and conditions that would change the decision.