AI's real threat at work: not job loss, but surveillance and control

The debate over artificial intelligence and jobs has missed the real danger unfolding in workplaces worldwide. While some warn of mass job losses and others celebrate productivity gains, neither tells the story happening now from Britain to Kenya to the United States.

For higher-paid workers in roles like analysis, law, consulting and management, AI acts as a helpful assistant. It removes routine tasks, speeds up decisions and creates room for creative thinking. These workers use AI to extend what they can do.

For millions of others, AI works differently. It does not assist them. It supervises them. AI appears in scheduling software that assigns shifts, route optimisation tools that decide how long tasks should take, and performance dashboards that measure whether workers are operating at full capacity. These workers are managed by AI, not supported by it.

This split is already visible. One-third of UK employers use monitoring technology to track workers' online activity. These surveillance systems show what comes next as AI spreads. Amazon software engineers report being watched and pressured to achieve more through AI, even when it slows their work. Meta plans to track employees' keystrokes, mouse movements and clicks to train its artificial intelligence.

Research on worker-AI interaction, cited in the 2024 White House economic report, shows the main issue is not immediate unemployment. It is the widening gap in skills, autonomy and wellbeing between those working with AI and those managed by it. Many jobs will exist in the future, but more will be pressured, fragmented and less human.

Work affects more than income. It shapes dignity, trust and control. During the pandemic, people understood how deeply work influences mental health. AI-managed workplaces intensify work pressure. When every click, step, call or pause can be measured and scored by a system workers cannot fully understand or challenge, stress follows. In warehousing, retail, hospitality, logistics and customer service, this means harder pressure from systems presented as neutral and objective, even when they are not.

Britain promotes itself as ambitious on AI. Plans exist to expand AI skills across the workforce. But organisations remain poorly prepared to introduce AI fairly. A global survey of business leaders found most say AI skills are competitive advantage, yet few budget meaningfully to develop employee skills. Even fewer have strong oversight. Many managers lack real responsibility for helping teams adapt. This is how inequality hardens.

If higher-paid workers receive training to use AI while lower-paid workers simply face surveillance and automated management, progress will not be shared. It will deepen imbalance. Workers need meaningful training in digital tools and broader skills: judgment, communication and critical thinking. They need democratic principles in the workplace. Systems affecting pay and performance should be transparent and open to challenge. Workers need a voice in how these technologies arrive.

AI introduced behind closed doors and justified through efficiency language ignores the people it affects most. Research shows involving workers improves job quality and helps employers integrate AI more effectively. The choice about how AI reshapes work is not made in Silicon Valley or at summits. It is made now, workplace by workplace, across Britain and globally. Unless attention turns here, the AI divide will become another quiet inequality that embeds itself before anyone notices it is everywhere.