AI Adoption Creates New Layer of Workplace Health Risks and Opportunities

As artificial intelligence reshapes workplaces across every sector, it's creating a catch which is dangerous to ignore. The technology designed to reduce workload is generating entirely new occupational health risks. Role ambiguity and hidden supervisory burdens are just the short-term issues. There are long-term concerns, such as erosion of critical thinking skills and the psychological demands on workers transitioning from task execution to AI stewardship.

As health and safety professionals, we’re at the cutting edge of this change and must be proactive in how we handle it.

What Are the Emerging Occupational Health Risks from AI?

Artificial intelligence is creating a new layer of occupational health risks that many organisations aren't prepared for. As workers transition from executing tasks to stewarding AI outputs, we're seeing novel psychological demands, role ambiguity, and hidden workloads that require immediate attention.

I've spent nearly four decades in health and safety, and this feels like one of the most significant changes I’ve seen in recent years. Although it comes with both positives and negatives, its completely changing what businesses expect from human workers.

The Shift from Execution to Overseeing

Workers who controlled specific tools for their roles are now managing AI systems across entire workflows, briefing AI agents, reviewing outputs, and taking accountability for results they didn't directly produce.

A recent editorial from Imperial College London highlights this brilliantly. As Lara Shemtob notes, this shift "requires supervisory skills and has far-reaching implications for health and performance at work."

Occupational Health and Safety Concerns

• Role ambiguity – Uncertainty about responsibilities when AI handles routine tasks.
• Hidden workloads – AI supervision demands aren't being quantified or resourced.
• Skill degradation – Weakened ability to respond to emergencies.
• Cognitive strain – Managing multiple AI agents whilst maintaining oversight.
• Critical thinking erosion – Research shows these decreases when workers have high trust in AI or low confidence in their own abilities.

I've seen this in safety management. When automated risk assessment tools or systems we’ve put in place suggest controls, there's real temptation to accept them without challenge.

AI doesn't understand context as well as an experienced professional does. It doesn't know about the machine acting up, the contractor taking shortcuts, or production pressures that might compromise implementation. This is where we need to think carefully about how we use it.

Why Health and Safety Professionals Must Get Involved

According to 2024 data, 62% of high-adoption organisations involve HR in AI strategy. But where's occupational health in that conversation?

We possess unique expertise in the relationship between work, health, and productivity. As AI reshapes conditions across sectors, from construction to healthcare, from manufacturing to professional services, our role is to protect the health of our workers throughout this change.

Practical Actions to Take Now

1. Conduct AI-specific risk assessments Traditional DSE or workload assessments won't cut it. Ask: How much time reviewing AI outputs? What happens when AI produces errors? Can workers challenge AI recommendations confidently?
2. Establish role clarity Define explicitly what decisions AI makes autonomously, where human judgment is required, and accountability structures when things go wrong.
3. Build AI literacy Workers need to understand AI's capabilities, limitations, how to spot errors, and when to escalate concerns.
4. Monitor for hidden workloads Track whether promised efficiency gains materialise or whether workers simply absorb additional supervisory duties without resources.
5. Protect cognitive capacity Ensure workers maintain emergency response skills. Consider simulation training and competency maintenance programmes.

The Opportunity Alongside Risk

AI offers significant opportunities when we act with intention. Tools can reduce barriers for neurodivergent workers through transcription and scheduling support. Digital assistants aid those with executive function difficulties.

The difference between AI as risk and opportunity comes down to implementation: involving the workforce, assessing health impacts, building safeguards, and maintaining human agency where it matters.

Your Role in This Transformation

The pace of AI adoption is high, even in historically slow-changing sectors. We can't assume someone else is managing the human factors.

Through Astutis’ consultancy work, I’ve seen successful organisations treat this as change management with significant health and safety implications.

Whether you're an OSH professional, safety manager, or business leader responsible for workplace health, now is the time to get involved in your organisation's AI strategy. This transformation needs your expertise.

Stay ahead of emerging workplace health and safety trends. Sign up for the Astutis Quarterly Newsletter for expert insights delivered directly to your inbox every quarter, or explore more articles and sign-up to our This Week in Health and Safety series below.