The 'robot relations' department may become reality in workplace of the future

When the first employee walks into a company's human resources department to complain about a robot co-worker or overzealous AI scheduling assistant, what exactly is HR supposed to do? Maybe refer the worker to "Robot Relations." The idea may sound like science fiction, but it is an idea emerging in one form or another during discussions about AI's impact on jobs, with the intention of getting ahead of a not-to-far-into-the-future workplace problem. "In the emerging AI-based world, employee concerns no longer will center on fellow workers but on interactions with AI agents, cobots, and large language models," Darrell M. West, senior fellow at Brookings Institution, wrote in a July commentary. Humans will be interacting with a variety of AI technologies, from assistants to agents and chatbots, and in the physical world (e.g. in a manufacturing space), cobots that may become core partners in work relationships. West wrote that if you can imagine all the ways in which tensions between human employees have to be resolved by HR, map that onto human resistance to technological change, and you can get a taste of what's coming. And with most existing employee support systems derived from an industrial era ill-equipped to deal with the grievances, challenges, and anxieties of the AI era, "businesses will need to supplement HR with robot relations (RR)," West wrote. West may be one of the only – if not the only – person to have so far used the "RR" term specifically, but what he is getting at is a big topic of conversation in HR circles. Complaints human workers will have about AI Expect the complaints from human workers to be varied, and potentially difficult to address: Can I be forced to use AI in my job if I resist? Whose fault was the mistake made on a project: the human worker's or the AI's? Human-AI tensions in the workplace already exist, with allegations that algorithms being used by HR today are unfair or biased, and in some cases, may be used to identify employees for layoffs. A bill recently passed by California's legislature with an overwhelming majority would ban the use of "robo bosses" for firing and/or disciplining workers. It awaits Gov. Gavin Newsom's signature — he has until the end of September to sign or veto it. Newsom vetoed an earlier version of the bill last October saying changes were needed, which sponsors say the new bill reflects. At an experimental San Francisco store operated by AI company Andon Labs, an AI system called Luna was given responsibility for running substantial portions of the business, including hiring, scheduling and managing workers. In August, Luna made its first recommendation to terminate an employee. The attention generated by this AI management experiment was among the reasons California lawmakers moved to restrict automated workplace decisions. Meta is facing a lawsuit, filed in July, in which former employees allege AI-assisted systems were used to rank and select workers for layoffs, disproportionately affecting employees who had taken medical or family leave. Meta has denied the allegations. Walmart and Amazon workers are increasingly concerned that HR decisions are being automated, according to a May survey released by nonprofit United for Respect, which has its roots in the labor movement and retail sector specifically. Amazon said the survey represents a fraction of a workforce in the millions. In June, Walmart shareholders rejected a proposal from the nonprofit that would have required more reporting on workplace automation and the impact on the workforce, which Walmart's management says it already provides transparency on. AI governance nonprofits say corporations have much more they should consider disclosing as the AI era develops. According to Just Capital, which studies business leadership including through analysis of corporate disclosures, half of the largest companies in the U.S. are not disclosing on the AI issues ranked as most important by the American public, among them, keeping humans in charge or at least in the loop. In a recent review of over 1,000 companies' disclosures on AI governance, data protection, and workforce impacts – which included a review of the eight tech sector "hyperscalers" – most were found to be limited in detail, according to Just Capital. The International Labour Organization, which defines algorithmic management as systems that organize, assign, monitor, supervise and evaluate work – as opposed to human managers doing all of those tasks – says use of these systems is growing steadily across industries ranging from customer service to transportation and logistics, banking, and health care. ILO said in a recent report that while AI systems' recommendations and real-time information can "enhance professional judgement," in other cases they mandate specific workflows, standardize procedures, and reduce human employee discretion. "These changes affect job satisfaction, skill use and the meaning workers derive from their work," it stated. The ILO expects these dynamics to spread across a greater number of high-level jobs previously defined by autonomy, and it sees additional reasons for workers to be concerned. "The integration of AI into work processes is often accompanied by expanded data collection and monitoring. AI systems can track outputs, behaviours and work rhythms in real time, enabling new forms of performance measurement. While this may improve coordination, it also raises concerns about digitally driven surveillance and information asymmetries between employers and workers. Workers may not always be aware of what data are collected or how they are used, particularly where regulatory safeguards are weak," ILO wrote. Survey results from the OECD published late last year found that algorithmic management software is being widely adopted across countries, led by the U.S., where 90% of managers say their firms have adopted at least one tool to "instruct, monitor or evaluate workers." Unions already fighting AI battles, but many workers lack protections Unions representing workers are actively engaging on the AI issue in collective bargaining that cover many core areas of labor interest and history: retraining, new job opportunities, and adjustments to compensation models so employees share in the productivity from new technologies, as well as severance pay and early retirement incentives, according to an MIT review of the early days of the generative AI era. And more specific AI protections are being sought, including over the use of personal data and use of AI-generated data specifically for employee discipline. "These will be cutting-edge issues in collective bargaining negotiations across a wide array of companies and industries," the MIT researchers wrote. For workers in the private sector lacking union representation, a department within the corporation may be the only option. "I think the idea of having such a department – or of existing HR departments expanding their remit in this direction – is worth taking seriously," said Orestis Papakyriakopoulos, Technical University of Munich professor of societal computing, which is an interdisciplinary field focused on how new technology impacts human society at the same time that it is being shaped by it. He says RR could address questions such as how AI agents and AI-based systems affect workers' well-being, empowerment, professional growth, and career trajectories, "and when they may instead create barriers to progression." "There is a need to assess how AI is changing what we value in the workplace," Papakyriakopoulos said. "We need to think about how to protect employees, reskill and retrain them, and clearly define the kinds of impact and contributions that should be valued in AI-mediated workplaces." Papakyriakopoulos said his team has already encountered tense dynamics on projects where it has been asked to develop AI systems for workplaces, with employees proving to be reluctant to participate. Workers have intentionally withheld information, provided inaccurate information, or otherwise disrupted the process because they were concerned about their position in the company and what the introduction of AI might mean for their job security, Papakyriakopoulos said. "Employees may already perceive AI deployment as something that can affect their power, status, and employment prospects," he said, and that makes an RR department an organizational idea that is needed right now. Other experts say teaching humans how to interact with machines will be required, but is only a subset of the issues that will be raised. Society will have to decide what happens when the interests of the humans and the machines' owners diverge. "Many so-called AI-driven systems are designed to reduce workers to mere extensions of the machinery, employees are often left to supervise automated workflows, resolve malfunctions, troubleshoot delicate equipment, and step in whenever a machine fails," said Cristina Zaga, assistant professor at the University of Twente. "Workers end up supporting automation's limitations," she added. Zaga says that as we confront new workers-machine relationships, the thinking cannot stop at the "new colleague" but extend to the new "means of production … owned and governed by corporations." She argues that any new organizational department that is properly conceived will not be designed with the overarching goal of keeping humans happy with machines, but with the goal of giving workers a meaningful role in the governance and design of automation. That may not be called "robot relations," but it could mean that a "workers-automation relations" department – or whatever you prefer to call it – will be part of the office of the future.