Chloe Maraina, our resident Business Intelligence expert, brings a unique perspective to the high-stakes world of AI integration within the modern workforce. With a profound aptitude for data science and a clear vision for the future of data management, she bridges the critical gap between technical innovation and the complex legal frameworks organizations must navigate today. Following her recent insights shared at the 2026 annual meeting of the Society for Human Resource Management (SHRM), Chloe provides a deep dive into the shifting landscape of employment law and technological ethics. She outlines the necessity of a collaborative framework involving HR, legal counsel, and IT departments to mitigate the multifaceted risks that arise when algorithms begin to influence human careers. This discussion explores the precarious balance between efficiency and liability, the psychological impact of AI on employee trust, and the strategic decisions companies must make in an era of evolving regulations.
We explore the primary legal challenges facing organizations today, focusing on how automated systems can inadvertently introduce bias and the specific steps leaders must take to ensure their data remains secure and their processes remain defensible. Chloe also addresses the “trust deficit” that often plagues AI rollouts and provides a roadmap for maintaining transparency and fairness in high-pressure scenarios like mass layoffs or global technological deployments.
When automated systems are allowed to completely replace human decision-making in critical areas like hiring or performance management, why do litigation risks regarding bias and discrimination increase so dramatically?
The moment you remove the human component from a decision-making chain, you fundamentally raise the bar for your legal defense. In the United States, we are currently operating within a complex patchwork of laws where AI implicates dozens of different regulations across various jurisdictions, making a singular compliance strategy nearly impossible. Trying to defend an automated process that runs on its own is significantly harder than defending a human-led process where specific directives were given and documented risk assessments were conducted. When a candidate or employee feels they were adversely impacted by a tool that ignored their objective qualifications, the lack of human oversight makes the organization look reactive rather than proactive. We see litigation risks spike because these tools can have inherent biases that, once deployed, create a paper trail of automated decisions that are difficult to justify in court without a human who can explain the “why” behind the output.
Beyond the well-known issues of bias and discrimination, what are the secondary risks to a company’s information and intellectual property that often catch organizations off guard during an AI rollout?
There is a very real danger regarding the leakage of trade secrets and confidential information, especially when employees turn to public AI tools because the enterprise hasn’t provided its own secure framework. We have already seen instances where sensitive company data was used to train another company’s model, effectively mixing proprietary information into a public pool where it can no longer be protected. Furthermore, the intellectual property risks escalate when contractors are brought in to develop custom AI models; if a contractor’s data is used for training, they might later claim IP rights over the entire model itself. This lack of clear documentation and specific contract language creates a legal vacuum that can lead to long-term disputes over who actually owns the intelligence the system has generated. Organizations often fail to realize that their most valuable data assets are being used as fuel for these engines, and without strict guardrails, that fuel can easily end up in the hands of competitors or third-party vendors.
Given that risks can vary significantly depending on whether a company is a small nonprofit or a large publicly traded corporation, which specific areas should HR leaders prioritize to protect their organizations?
The absolute first priority for any leader is to identify and address any automated system that completely bypasses human intervention, as ensuring a “human-in-the-loop” is the strongest defense against a legal claim. Second, it is vital to ensure that HR is actually in the room when these tools are being selected and adopted, rather than letting the IT department drive the process in a silo. We have seen tech-heavy rollouts go south very quickly because the innovators didn’t account for the real-world implications of workforce deployment, such as the required disclosures and compliance communications. This collaboration between legal counsel, IT, and HR is essential because an AI rollout that looks like a simple employment issue can quickly morph into an unfair trade practice or an intellectual property dispute. When HR is excluded, the company loses its first line of defense in explaining the technology to the people it affects, leading to a breakdown in both compliance and corporate culture.
With the regulatory environment at the federal, state, and international levels constantly shifting, is it more prudent for organizations to take a wait-and-see approach or should they strive for immediate, total compliance?
Total, 100% compliance with every global AI and privacy law is effectively impossible right now, much like the challenges we have seen with GDPR and various U.S. state laws. You have to pick your battles and focus on complying with the specific pieces that carry the most significant weight in your primary jurisdictions while performing a constant risk-benefit analysis. For example, some companies in California are forced to decide whether to comply with a Data Subject Access Request (DSAR) under the CCPA or risk giving a potential litigant free discovery through the disclosure process. There are inherent risks in both over-disclosure and under-disclosure, so being a “trailblazer” who implements every new tech trend can actually paint a target on your back. Often, it is wiser to let the unsettled areas of law stabilize before pouring millions into a full-scale AI implementation that might be deemed non-compliant by a new statute six months later.
You’ve mentioned that HR’s role has shifted from being a reactive gatekeeper to being the first line of defense regarding employee reactions; how does the concept of “trust” factor into a successful AI integration?
Trust is the single most important factor in how employees react to new technology, and HR is the department responsible for building and maintaining that foundation. When trust exists, an employee is more likely to cooperate with the company, much like how they might willingly hand over a personal phone for a forensic investigation because they believe the organization won’t abuse their privacy. If there is a “trust deficit” within the company—perhaps due to existing issues with pay equity or unfair decision-making—then an AI rollout will be viewed with immediate suspicion and fear. HR must be able to explain not just what the tool does, but why it is being used and how it benefits the individual worker, which requires HR leaders to be fully convinced of the technology’s value themselves. When communication lines are fractured, every automated output is scrutinized as a potential threat, whereas a high-trust environment allows for a much smoother transition and more effective reskilling efforts.
If an organization is facing a reduction in force (RIF) and employees suspect that AI was used to select individuals for layoffs, what are the specific legal landmines that the company needs to avoid?
Selective layoffs carry a much higher litigation risk than a total plant closing because the core question becomes whether the AI tool targeted people in a protected category. For instance, if a company has two departments performing similar tasks and decides to deploy AI-driven terminations in only one of those departments, and that department happens to have a higher percentage of women, you have a clear case for a discrimination claim. Litigation risk is almost guaranteed if AI implementation is not uniform across the board or if the criteria the algorithm used to select individuals cannot be objectively explained. Employees will naturally ask why they were chosen for replacement while others were not, and if the answer is buried in a “black box” algorithm, the company will struggle to satisfy the requirements of the Worker Adjustment and Retraining Notification (WARN) notices and other legal protections. Every case is unique, but the common thread is that any lack of uniformity in how AI is applied during a layoff provides fertile ground for legal challenges.
What is your forecast for the federal and state regulatory landscape regarding AI in the workforce?
At the federal level, I expect we will see a continued hands-off approach for some time, although existing protections like Title VII against discrimination will remain the primary tools for enforcement. We are already seeing a rise in litigation based on the theory that AI tools scraping social media or the internet for background checks should be bound by the Fair Credit Reporting Act (FCRA), even if the federal government hasn’t yet issued a definitive ruling. While the federal government has historically stepped away from comprehensive privacy laws, leaving the heavy lifting to the states, we are seeing a massive uptick in pending state legislation and privacy statutes. Even without a central federal mandate, companies will find themselves in a very restrictive environment because the combination of state laws and aggressive private litigation will effectively force them into the same high-compliance standards. Organizations should prepare for a future where state-level privacy rights dictate their national AI strategies, as the risk of facing fifty different sets of rules will eventually demand a “highest common denominator” approach to policy.
