Should Your Company Have a Robot Relations Department?
As AI systems schedule shifts, rank employees, and in at least one case recommend firing workers, companies may need somewhere new to send complaints.
Key points
- Darrell M. West, a senior fellow at the Brookings Institution, proposed "robot relations" (RR) in a July 2026 commentary as a formal workplace function to handle human-versus-AI conflicts.
- A California bill passed with an overwhelming legislative majority in 2026 would ban AI systems from making firing or disciplinary decisions, and awaits Governor Gavin Newsom's signature by the end of September.
- At a San Francisco store run by AI company Andon Labs, an AI system called Luna made its first recommendation to terminate an employee in August 2026.
- Meta faces a lawsuit filed in July 2026 alleging AI-assisted ranking tools selected workers for layoffs, disproportionately affecting those who had taken medical or family leave.
- 90% of U.S. Managers say their company has adopted at least one tool to instruct, monitor, or evaluate workers, according to OECD survey data published in late 2025.
AI software is now scheduling shifts, grading performance, and in a growing number of workplaces, deciding whether workers keep their jobs. HR departments weren't built for any of that. When an employee wants to push back against an algorithm, there's often nobody to call.
Darrell M. West, a senior fellow at the Brookings Institution, proposed one answer in a July 2026 commentary: a dedicated "robot relations" department, or RR. Think of it as HR, but instead of mediating disputes between employees, it mediates disputes between employees and the AI systems they work alongside. CNBC Tech first reported on the idea gaining traction in HR circles.
Why is this becoming urgent now?
The cases stacking up are hard to ignore. At an experimental San Francisco shop run by AI company Andon Labs, a system called Luna managed hiring, scheduling, and worker oversight. In August 2026, Luna recommended firing an employee, the first known public case of an AI system directly flagging a termination. That story was among the reasons California legislators moved to restrict automated workplace decisions.
Meta faces a July 2026 lawsuit in which former employees claim AI-assisted ranking tools selected workers for layoffs, hitting people who had taken medical or family leave hardest. Meta denies the allegations.
A May 2026 survey by nonprofit United for Respect found Walmart and Amazon workers are increasingly worried about automated HR decisions. The International Labour Organization warns that AI monitoring can track workers' outputs and behaviours in real time, and that workers often don't know what's being collected or how it's used.
Our reporting on 19 September found that most people worldwide already fear AI will cost them their jobs, drawing on a Pew poll of more than 42,000 people across 37 countries. What a robot relations function would add is somewhere to go when those fears become specific, documented grievances.
What would 'robot relations' actually do?
It'd answer questions existing HR can't. Can your employer force you to use an AI tool you distrust? If an AI system makes a mistake on a project, who's accountable? How does a worker appeal a decision made by an algorithm?
Orestis Papakyriakopoulos, a professor of societal computing at the Technical University of Munich (societal computing is the study of how technology reshapes society), told CNBC Tech the idea is worth taking seriously. He frames it around worker career development and the risk that AI systems create invisible barriers to promotion.
Union members in workplaces with collective bargaining, meaning formal contracts between workers and employers, are already pushing for AI protections: retraining rights, limits on personal data use, a share of productivity gains AI creates, and stronger severance provisions, according to an MIT review of early generative AI negotiations. Workers without union representation have far fewer options.
A California law requiring human sign-off on AI-driven firings and discipline, if Newsom signs it, would be a first step. Newsom vetoed an earlier version last October. Legislation moves slowly; AI adoption doesn't.
Should you worry about this now?
Probably yes, and sooner than most people expect. The Andon Labs experiment and the Meta lawsuit aren't edge cases. They're early examples of a dynamic that the OECD data confirms is already mainstream: nine in ten U.S. Managers report using at least one tool that instructs, monitors, or evaluates the people who work for them.
The practical gap right now is accountability. Most workers have no clear channel to challenge an automated decision, and most employers haven't built one.
What to watch for: If your employer uses software to schedule your work or flag performance issues, ask HR in writing what data the system collects and how a human reviews its decisions. In most jurisdictions, you have a right to know. If the answer is vague, document it.
