Moving fast with generative artificial intelligence (GAI) without thinking critically about its implications could cause downstream problems and potential liability. For law firms, focus must be given to the next generation of attorneys, who will likely use and rely more on GAI in their practice, and steps should be taken to ensure they continue to develop the analytical skills and professional judgment that define great lawyering. For labor and employment litigation, GAI has opened the courthouse doors to a surge of pro se plaintiffs while raising novel legal questions about work product, privilege, and discoverability that courts are only beginning to sort out. On the other hand, many employers are rushing to deploy GAI-powered tools in the workplace in areas such as hiring and discipline, where real challenges may exist in the form of algorithmic bias that could result in unintended unlawful employment practices if not appropriately assessed and managed.

GAI is a powerful tool, but deploying it effectively requires more than enthusiasm. It requires intention, oversight, and a willingness to grapple with hard questions.

GAI IN LAW FIRMS: GAI AND ASSOCIATE GROWTH

Before addressing problems that employers might face with GAI, it is worth evaluating the potential challenges GAI poses within law firms. Many tech-savvy associates now use GAI tools to perform the initial legwork for many assignments. For example, upon receiving a considerable volume of documents, an associate might load those documents into a GAI tool to identify what these documents consist of, generate a timeline, and explain how these documents support or defend the claims in a case. When dealing with a discovery motion, an associate might load the brief into a GAI tool and ask it to prepare a draft response. When presented with an opportunity to handle a complex legal issue, an associate might turn to a GAI tool for guidance on where to start. Prior to the advent of GAI, these were the types of tasks that associates were expected to perform through critical thinking, consultation with more senior attorneys, and/or manual research on their own. While more timeconsuming and sometimes extremely tedious, these assignments were also valuable learning tools and opportunities for associates to hone their legal skills, develop their legal reasoning skills and acumen, and eventually become proficient in their chosen profession.

Historically, associates developed the necessary legal skills, such as analytical reasoning, anticipating opposing counsel’s arguments, formulating practical solutions, and tailoring advice to clients’ needs (i.e., “soft” or “EQ” skills),¹ by performing the fundamental and necessary legwork (i.e., “hard” skills like research, writing, reviewing, and categorizing documents). With GAI, associates are now able to partially or entirely bypass this crucial stage of professional development.

Nonetheless, the reality of the situation is clear. Clients often look to their counsel to handle their needs with financial efficiency—and utilizing GAI to perform the legwork is one of those measures that align with the clients’ expectations. This means that early career attorneys will now need to develop their EQ skills another way, and it will likely take strong support from more experienced and established attorneys.

Forward-thinking firms are already evaluating and addressing these issues. It should come as no surprise that associates who are included in client meetings and conversations with leadership teams — not sporadically but regularly — feel more valued and supported. Inclusion in client meetings provides associates with valuable insight into a client’s operations and challenges as well as into how the leadership interacts with the firm’s clients, and with opportunities to observe, firsthand, how more experienced attorneys communicate their advice to the client. Increased collaboration is also a tool to instill value for associates and develop their fundamental legal skills. With GAI tools supplementing some of the more traditional ways to develop legal skills for associates in their formative years, firms may need to devote more time to nurturing their associates through collaboration and a more inclusive approach to their client relationship.

GAI IN LITIGATION: GAI IS ALLOWING CLAIMS THAT WERE PREVIOUSLY REJECTED OUTRIGHT TO PROCESS

The proliferation of GAI has ushered in a new era of pro se litigation that is concerning to most employers. Until recently, the complex procedural requirements for legal filings served as a barrier to many pro se claims without a legally cognizable cause of action. GAI has effectively dismantled that barrier. Pro se plaintiffs who previously would have been deterred by the intricacies of civil procedure, legal research, and persuasive legal writing are now armed with tools capable of generating complaints that satisfy the requisite pleading standards and motion papers accompanied by legal arguments. The result is an unmistakable uptick in self-represented litigants filing employment-related claims — many of which, in a pre-GAI world, would likely not have made it past the initial consultation with and assessment by a plaintiff’s side attorney.

According to Lex Machina’s 2026 Employment Litigation Report, federal employment lawsuits with unrepresented plaintiffs grew each year from 2021 through 2025.² In 2025, more than 16 percent of federal employment lawsuits were filed by individual plaintiffs without legal representation.³ Courts across the country are grappling with dockets swelling with pro se filings that bear the unmistakable fingerprints of GAI assistance.⁴ While these filings may not equate to pleadings prepared by licensed attorneys, they nevertheless articulate a cognizable legal theory such that Courts would decline to dismiss these claims outright. For employers, this means that claims with marginal merit that might have been dismissed at the initial pleading stage now require substantive legal defense.

Perhaps more concerning are the novel legal questions emerging from this trend. Courts have begun to wrestle with whether pro se plaintiffs can assert work-product protection over their GAI tool use, such as their queries and outputs. Recent decisions have reached strikingly different conclusions.

In United States v. Heppner, the Southern District of New York answered that question with a resounding no — at least on the facts presented.⁵ In Heppner, a criminal defendant under indictment for securities fraud used the publicly available AI platform Claude to independently prepare documents outlining potential defense strategies and legal theories. He later asserted both attorney-client privilege and work-product protection over those AI-generated materials. The court rejected both claims. On the work-product question, the court held that the doctrine “provides qualified protection for materials prepared by or at the behest of counsel in anticipation of litigation,” and that Heppner’s AI documents failed that test because they “were prepared by the defendant on his own volition” — without any direction from counsel.⁶ The court further noted that while the documents may have influenced defense counsel’s strategy going forward, they did not reflect the counsel’s strategy at the time they were created. In short, because no attorney directed Heppner to use Claude, and the documents did not expose any attorney’s mental processes, the work-product doctrine offered no shelter to the unrepresented, individual criminal defendant.

Closer to home, the Eastern District of Michigan in Warner v. Gilbarco, Inc.⁷ reached a markedly different outcome. In Warner, a pro se employment discrimination plaintiff used ChatGPT extensively in connection with her litigation. Defendants moved to compel production of all documents and information concerning her use of AI tools, arguing that any work-product protection was waived when she disclosed her mental impressions to a third-party AI platform. The court disagreed and denied the motion. The court explained that “[a] pro se litigant has a right to assert work-product protection over such material” and held that using ChatGPT does not constitute a waiver of that protection, reasoning that the “work-product waiver has to be a waiver to an adversary or in a way likely to get in an adversary’s hand.” Critically, the court concluded that “ChatGPT (and other generative AI programs) are tools, not persons” — meaning disclosure to an AI platform is categorically different from disclosure to a human third party that would ordinarily trigger a waiver. The court characterized the defendants’ demand as “a fishing expedition” into the plaintiff’s “internal analysis and mental impressions” and warned that the defendants’ theory would “nullify work-product protection in nearly every modern drafting environment, a result no court has endorsed.”

Taken together, Heppner and Warner illustrate the rapidly evolving and unsettled nature of this area of law. What does this mean for employers and their HR teams? The answer, while perhaps unsatisfying in its simplicity, is that prevention must become the paramount focus. When the previously available procedural barriers to filing litigation have been lowered so dramatically, the only effective countermeasure is to reduce the occasions that give rise to claims in the first place. This requires a renewed commitment to compliance — not as a check-the-box exercise but as a living, breathing component of organizational culture. Management training has become more critical than ever. Supervisors and managers should be armed with tools to recognize potential legal landmines. This means regular, substantive training on topics including harassment prevention, reasonable accommodations, wage and hour compliance, and proper documentation practices.

Equally important is attention to employee morale. Employees who are dissatisfied with their workplace tend to become disgruntled and litigious. In an era where GAI can transform workplace grievances into valid legal action with relative ease, attention to the human elements becomes paramount. Exit interviews, frequent check-ins, anonymous feedback opportunities, and genuine responsiveness to employee concerns are no longer just best practices — they have now become necessary risk mitigation strategies.

Lastly, management teams should be cautious in using GAI tools themselves. While undeniable efficiencies provided by these GAI tools may be alluring, these GAI tools are not substitutes for human judgment, especially in matters that carry significant legal risk. As seen in Heppner, GAI searches and outputs that contain leadership’s internal thought processes that were not formulated at the direction of legal counsel could become discoverable if the matter proceeds to litigation.

GAI IN THE WORKPLACE: AUTOMATED DECISION-MAKING SYSTEMS STILL REQUIRE HUMAN OVERSIGHT

Beyond litigation, GAI has presented a host of challenges for employers who have integrated these tools into their day-to-day operations. Automated decision-making systems (ADMS) powered by GAI are being increasingly utilized in hiring, performance evaluation, promotion decisions, and workforce management. While these systems promise objectivity and efficiency, they also carry significant legal and ethical risks.

The idea of removing human subjectivity from employment decisions is appealing. After all, how can the decision be discriminatory if a cold, calculated machine makes the decision? The reasoning, however, is fundamentally flawed when one considers the fact that GAI systems are trained on historical data, and that data often reflects the very biases we seek to eliminate — commonly referred to as “algorithmic bias.”⁸

Changes in the federal administration have caused regulatory bodies to revise or retract previous AI-related guidance for employers. In 2025, the U.S. Equal Employment Opportunity Commission (EEOC) retracted guidance on AI in employment selection procedures.⁹ And the Department of Labor has placed a disclaimer on their framework for AI-powered recruitment tools, stating that it “may be out of date or not reflect current policies.”¹⁰ However, state and local governments have continued to advance their own AI-workplace legislation. Illinois now requires employers to notify candidates when AI is used in video interview analysis and recently amended their Human Rights Act to include broader prohibitions on the use of GAI in hiring, promotion, and discipline decisions that could result in discrimination.¹¹ New York City’s Local Law 144 mandates bias audits for automated employment decision tools.¹² Although Michigan has not yet enacted legislation around AI in the workplace, House Bill 5579 was introduced on February 24, 2026, that would ban employers from using AI programs to make decisions related to setting wages, hiring and firing workers, and tracking employees’ facial patterns.¹³ Under the proposed legislation, employers would also need to get written consent from employees when using an AI tool to monitor productivity.¹⁴

For employers seeking to harness the benefits of GAI while complying with applicable law and mitigating risks, several principles should guide implementation.

First, transparency is paramount. Employees and applicants should be informed when ADMS are being used and should understand, at least in general terms, how those systems operate.

Second, human oversight should be a central component of any ADMS deployment. These systems should be designed to augment, not replace, human decision-making. Critical employment decisions — terminations, denials of accommodation, promotion decisions — should always involve meaningful human review, from someone with the authority to override the system’s recommendation when appropriate.

Third, regular auditing is essential. Employers should periodically assess their ADMS for disparate impact and other discriminatory effects. This auditing should be conducted by qualified professionals and should examine outcomes across protected categories.

Fourth, the use of ADMS should be documented in detail. Employers should maintain records of what ADMS are in use, how they were developed or procured, what data was used to train them, and what oversight mechanisms are in place. This documentation will prove invaluable in the event of a legal challenge and demonstrates a good-faith effort to deploy these tools responsibly. Finally, employers should resist the temptation to view GAI as an “out” for difficult decisions. The most sensitive workplace matters — those involving discipline, accommodation, and termination — require the exercise of human judgment, empathy, and contextual understanding that GAI cannot replicate.

To conclude, GAI is here to stay. It is imperative for us in the legal profession to find a way to leverage its efficiencies without sacrificing the human judgment, mentorship, and compliance culture that remain essential to the practice of L&E law.