Could AI write your next pink slip? Sen. Bob Casey proposes legislation to keep humans in HR
BETHLEHEM, Pa. — For decades, American pop culture has flooded movie theaters with cautionary tales of robots turning on their human handlers and rogue AI. And while fears of SkyNet may not yet be on point, AI is already terminating American workers — from their jobs.
Artificial intelligence is still in its infancy, but major corporations are embracing the emerging field more and more. When the tight marketplace sent businesses scrambling to find employees, many of them adopted AI and automated processes to keep up. A handful have even taken to using algorithms to handle firings with little human oversight.
- Businesses are turning to AI and computer programs to assist with hiring, disciplining and firing employees
- The emerging technology is advancing quickly, but the U.S. has no legislation regulating its use
- Sen. Bob Casey has sponsored a bill that would require human oversight of these computer systems and testing to ensure they don't violate workers' civil rights
But the practice has emerged quickly with minimal oversight; the U.S. has no overarching laws governing artificial intelligence. Sen. Majority Leader Chuck Schumer, D-NY, said last month he wants to introduce legislation regulating AI later this year, but what that looks like remains unclear.
Sen. Bob Casey, D-Pa., is hoping to shape that debate in favor of employees. A bill he introduced last week, the "No Robot Bosses Act," would require human oversight of personnel decisions and add other regulations meant to protect workers' rights, he said in a news release.
“As robot bosses become more prevalent in the workplace, we have an obligation to protect working families from the dangers of employers misusing and abusing these novel technologies,” Casey said in a news release.
The release cited news reports to emphasize these are not just hypothetical situations. AI has been widely adopted by companies in an effort to boost hiring in the tight labor market, but those tools have the potential for serious oversights. The Equal Employment Opportunity Commission warned last year that these programs can fail to account for federal guidelines meant to protect people's civil rights. For example, an algorithm may rule out a qualified applicant due to a gap in their work history. But if that gap is due to the applicant taking leave to treat a disability or give birth, that amounts to discrimination under federal law.
"In this market, where there's such a shortage of available workers, relying solely on technology without some reasonable backup is somewhat irresponsible."Tina Hamilton, CEO of myHR Partner
Bloomberg has also reported that Amazon has fired drivers without human oversight, instead relying on computer analysis of their performance. Past employees said they were harshly graded for things beyond their control such as getting a flat tire or delays caused by navigating snowy country roads. At times, the automated emails they received failed to explain what was the cause of their termination, Bloomberg reported.
Casey's proposed legislation would make it illegal for companies to hire, discipline or fire employees using automated systems without human oversight. It would require companies to be upfront about how AI is being used to make personnel decisions, mandate that the programs be tested to ensure they comply with workers' civil rights and that companies train personnel about potential shortcomings of the AI programs. The bill would also create a Technology and Worker Protection Division inside the Department of Labor to enforce the legislation.
Casey was not available for comment. However, an aide in his office speaking on background said that while Amazon has become a poster child of using algorithms for managing employees, it's far from the only one. With AI so widespread, Casey recognizes the need to have protections in place for workers across the board, the aide said.
"It's actually common in a number of other sectors — workers in retail, even in healthcare — it really goes beyond that. As the technology is implemented, it will continue to spread," the aide said.
Tina Hamilton, CEO of myHR Partner in Allentown, agreed. She said that large and mid-sized businesses have already integrated AI and algorithms into their HR departments and the ones who haven't are falling behind. She's found the best uses for it are in roles that don't require emotional intelligence, such as overseeing payroll. Human judgment is still the best way to determine how qualified someone is for a job and if they'll be a good fit for your culture, she said.
"In this market, where there's such a shortage of available workers, relying solely on technology without some reasonable backup is somewhat irresponsible," Hamilton said.
But even those companies delegating important tasks to robots need to have humans on hand ready to pivot in. If something goes wrong with the payroll processing, a human will be needed to fix the situation and reassure workers who are rightfully anxious about making their next mortgage payment, she said. If they don't, employees will find work at companies that invest more in humans.
"Organizations have to decide what kind of culture they want to have. If their focus is straight up on bottom line and using technology to create efficiencies, well then, they'll reap the benefits of that," Hamilton said.
Casey's aide acknowledged that creating the Technology and Worker Protection Division may be a tough pitch in this divided Congress. Republicans control the House and want to slash spending. Increasing the size of the Department of Labor is likely to run into stiff headwinds as a result. However, the federal government lacks the infrastructure to oversee the emerging field of artificial intelligence, making the division a necessity, the aide said.
"That is a harder sell, obviously, but it's one that when we spoke to the experts and the stakeholder groups, they felt very strongly about," the aide said.