Imagine your supervisor tells you to delegate all your new tasks to a teen intern: the CEO’s kid. The intern is certainly clever and quick, but their work can have serious problems. You have to pick through everything they turn in to make sure it is correct. When the process goes well, the intern does save you time. But if you don’t catch their mistakes, you will be held responsible. Depending on your position, your mistake might even go viral.
This situation would create plenty of labor problems on its own. But many employees will relate that hypothetical as similar to how they're being asked to use AI. Managers and CEOs often require employees to use AI tools to increase productivity, but even as they do, the software can introduce new problems. Worse, AI and algorithms may introduce biases that cause real harm or even violate employment law.
Can Employers Require Workplace AI Use?
The immediate appeals of AI technology are cost savings and increased productivity. According to Forbes, one study indicates a 4% productivity increase for software developers using generative AI, and Anthropic forecasts a 1.8% annual productivity gain across the entire U.S. workforce. That gives employers natural incentives to require workers to incorporate AI tools into their workflows. However, not every field — and not every worker — is the same, and employees may have reasonable objections to their own AI use. Do those employees have legal options?
Attorney Steven Obiajulu notes that as he writes, there is not yet any case law “saying an employer may, or may not, fire or refuse to hire someone solely because that person will not use AI tools.” Even so, a case like that is inevitable. In at-will employment states (almost all of them), businesses can terminate employment for almost any reason, including refusing to do work as instructed with AI tools. But the law does not give total free rein to at-will employers. They still cannot fire someone if the cause for firing is unlawful or against public policy.
Possible Grounds for an Employee to Refuse AI Use
Federal anti-discrimination law forbids employers from discriminating against workers who are members of protected classes — including disability and religion. States may have their own employment laws that mirror federal requirements or provide more coverage.
Under the Americans with Disabilities Act, if a disabled worker is unable to use AI tools, they might be able to claim a “reasonable accommodation” from their employer to use previous methods or different software, so long as that does not create an “undue hardship” for the company. For example, the American Foundation for the Blind points out that an employer’s chosen software often excludes users with low vision.
A worker of faith might also have grounds to request a reasonable accommodation based on a sincerely held belief about AI. As Attorneys Matthew P. Gizzo and James M. Paul discuss, the EEOC won a case against a company that denied a fundamentalist Christian an alternative to a hand scanner, which he considered to be a “Mark of the Beast” associated with the Antichrist.
Though specifics vary by state, workplace law does not allow firings done in retaliation or for any other unlawful purpose. If, for example, a company is using AI to break the law or retaliate against others, the worker might be in a stronger position to refuse. However, this would be part of a complex and sensitive legal issue for the employee, who would need advice as soon as possible.
When AI Technology Enables Discrimination
Lawmakers and industry critics have increasingly raised concerns about how easily AI technology can discriminate on behalf of its employers. Because developers use existing writings and materials to train artificial intelligence, conscious and unconscious bias can end up reflected in AI answers and tools. As a U.N. reporter puts it: “Bias from the past leads to bias in the future.” Without careful oversight, AI tools and algorithms can easily reflect and reinforce preconceived notions as they determine whether to hire, fire, or discipline workers.
In New York City, Local Law 144 prohibits employers from using automated employment decision tools (AEDTs) for hiring or workplace decisions unless they have had a recent bias audit that provides an impact assessment. California also has civil rights regulations on their use, and Governor Gavin Newsom may soon decide to sign laws passed to govern the use of AI for firing decisions, surveillance of workers’ emotions, and more.
Currently, the federal government disfavors state-level AI regulations and prefers a business-friendly, majoritarian approach to labor laws and enforcement. However, this cannot stop an individual’s exercise of their civil rights, especially under state law. This area of law evolves every day, so employees should watch developments carefully and consult an attorney about any serious issues they encounter.
