AI is transforming the way we recruit, evaluate, develop, and engage people within organizations. For HR teams, this is exciting, because it presents efficiency, scale, and insight to processes that are often way too manual. But it also raises ethical questions that go to the heart of trust, fairness, dignity, and employee well-being. As we look ahead, HR leaders have to be asking themselves how we can use AI not just to optimize business outcomes, but to protect and uplift individuals?
Why Ethics Can’t Be an Afterthought
HR doesn’t just serve business metrics. We’re a steward of culture, of morale, of people’s lives. When AI tools make decisions about people, whether that’s screening applicants, ranking performance, or guiding promotion pipelines, there should be transparency. Employees need to know that fairness and respect are baked in.
It’s about preserving trust. If decisions are perceived as opaque, arbitrary, or biased, it erodes loyalty and psychological safety.
What California’s New Regulation Tells Us
Earlier this year, California finalized regulations under the Fair Employment and Housing Act (FEHA) dealing with automated decision systems (ADS), which take effect October 1.
Here’s what HR teams in California, and any organization with employees or applicants who live/work there, need to understand from an employee-first lens:
- Definition is broad. “Automated Decision System” includes not just fully automated tools, but anything that assists human decision-making around employment benefits (hiring, promotion, training, etc.). So even tools that give managers recommendations, or screen resumes, may be in scope. Inside Tech Law+1
- Anti-discrimination protections apply. Systems that discriminate, whether intentionally or by proxy, based on protected categories (such as race, gender, disability, age, etc.) are prohibited. HR teams have to be prepared to show that the tools they use have been tested for bias.
- Transparency, human oversight, and vendor accountability. Employees (or applicants) impacted by ADS-influenced decisions should be able to understand what sort of system is used, how decisions were made, and whether there was human review. Also, the law expands who is considered an “agent” of the employer, which can include vendors or third parties whose tools influence decisions. That means responsibility doesn’t vanish just because you outsourced part of the algorithm.
- Recordkeeping and documentation. Employers must retain automated-decision system data, including scoring outputs, datasets, audit findings, etc., for four years instead of two. This matters a lot when defending against claims or doing internal continuous improvement.
- Proactive bias testing matters. Employers who never audited their AI tools for bias (or did so a long time ago) may be vulnerable. The quality, recency, scope, and results of bias testing (and the response to its findings) are now very relevant in defending decisions under FEHA.
And businesses outside of California may not think they need to worry about this. But if you have a remote or hybrid workforce, and may be recruiting someone in California, then this is definitely something you need to be aware of.
Ethical Risks HR Must
Putting the employee first means recognizing the risks:
- Reinforcing systemic bias. If your training data reflects past discrimination, AI can replicate it in subtle ways (e.g. favoring certain schools, favoring resumes with certain wording associated with specific demographics).
- Loss of agency and voice. Employees or applicants may feel they are being evaluated by an algorithm they can’t understand, challenge, or influence.
- Privacy intrusion. Tools that analyze video, voice, biometric or behavioral data raise sensitive privacy issues—how the data is collected, stored, used, and who has access.
- Unequal impact on vulnerable groups. People with disabilities, neurodivergent individuals, older workers, non-native speakers, people from marginalized communities may be disproportionately harmed if systems aren’t designed to accommodate diversity.
Employee-First Best
How can business lead with ethics while still reaping the benefits of AI?
We have to begin auditing all AI/ADS tools for bias now. Start with the tools that impact decisions related to hiring, promotion, and performance reviews. Ask vendors for their bias test reports, and commit to working with vendors who are making this a priority.
Be transparent with employees and candidates. Let them know that AI is used in the process. But make sure that they know that they have the option for meaningful human oversight. Employees need a chance to appeal or ask for reconsideration.
Minimize the collection of sensitive data. If such data is used, ensure strong protections, consent, and clarity about storage and use.
Ensure any AI systems accommodate different backgrounds. This goes beyond performative DEI policies. We have to take into account language, disability, and culture. Any system or process we use should be tested with diverse data sets and subject matter experts.
Putting It All Together
For HR leaders, the ethical imperative is clear: AI can make your work smarter, but only if used with care. In California, new laws make that care a legal requirement. For employees, the difference between an organization that handles AI poorly and one that prioritizes ethics can be profound/
Adopting AI with an employee-first mindset isn’t just morally right; it’s smart. It builds trust, reduces risk, improves retention, and enhances employer brand. When people believe they’re treated fairly, they engage more deeply, innovate more freely, and stay longer.
How is your organization ensuring AI decisions are fair, transparent, and centered on employees, not just efficiency? Need help unraveling it? Book a discovery call and let’s see what we can do together.
