What if the most useful thing an algorithm did for your promotion process was to tell a manager, "You may have overlooked someone." That is the pattern emerging from a new HiBob report, which found that more than half of HR managers using artificial intelligence to evaluate workforce decisions went back and reconsidered employees they had initially passed over for promotion. The finding is encouraging, but it raises harder questions. Can AI genuinely reduce manager bias, or does it simply shift where bias hides? And how should HR leaders think about those tools while the legal ground beneath diversity programs is moving?
The Promise of AI as a Bias Check on Promotion Decisions
Promotion decisions have long been shaped by familiarity, proximity, and informal networks. Managers tend to advance people they know well, which often means people who resemble them or who work closest to them. AI-assisted evaluation can interrupt that pattern by surfacing data that a busy manager might never revisit. According to the week's industry roundup, the HiBob findings suggest that this kind of prompt can change outcomes.
More than half of HR managers using AI to evaluate workforce decisions reconsidered employees they had overlooked for promotion.
What the Finding Does and Does Not Prove
It is important to read this result carefully. Reconsideration is not the same as promotion, and a manager who revisits a file has not necessarily made a fairer decision. The finding shows that AI can function as a prompt for review, not as an automatic corrective. The quality of the outcome still depends on the criteria the system uses, the data it draws from, and whether managers apply it with discipline.
That distinction matters because AI systems inherit the assumptions embedded in their training data and in the performance records they analyze. If past reviews reflected biased judgments, a model trained on those reviews can reproduce them with a veneer of objectivity. HR teams should treat AI-supported promotion tools as one input among several, and they should audit both the recommendations and the human decisions that follow.
Candidates Are Using AI to Judge Employers Too
The same technology reshaping internal decisions is changing how external talent evaluates employers. New research from Built In, described in a report on applicant behavior, finds that job seekers increasingly use generative AI tools to investigate potential employers. Over half of the candidates surveyed said they had chosen not to apply to certain companies based on what the AI told them.
Over half of surveyed candidates chose not to apply to certain companies based on insights AI provided.
This creates a two-sided problem. Internally, AI may help managers reconsider who deserves a promotion. Externally, AI may quietly filter applicants out before any recruiter sees their résumé. An employer's reputation, pay transparency, review practices, and public statements are now being summarized and scored by tools that candidates trust. HR leaders who rely on employer brand campaigns alone may be missing the conversations that actually determine whether qualified people apply.
Return-to-Office Mandates Meet the Collaboration Question
Another piece of the week's coverage challenges a common assumption behind return-to-office policies. A study by the Federal Reserve Bank of New York found that full-time workers spend roughly half of their in-office days doing solo work or joining video meetings. In other words, the office does not automatically produce the spontaneous collaboration that mandates are often meant to create.
For HR, this finding should prompt a practical question: what specific work is the office supposed to support? Mandates justified by collaboration will struggle if employees are spending much of their time on tasks that could happen anywhere. A more defensible approach links attendance to defined purposes, such as onboarding, cross-team design sessions, or mentoring, and then measures whether those purposes are being met.
The Legal Landscape Around Diversity Programs Is Shifting
The primary source also notes that discussions at the ABA Labor & Employment Law conference addressed how ongoing legal challenges are reshaping corporate diversity and inclusion programs. That reshaping is visible in recent enforcement activity. A $29.5 million settlement reached by Lockheed Martin with federal authorities over allegations that its inclusion and diversity practices discriminated against white employees signals that diversity initiatives are subject to the same anti-discrimination scrutiny as any other employment practice.
Employment attorneys are also tracking a different trend. As reported in a warning about religious discrimination claims, U.S. employers should expect more such charges as the Equal Employment Opportunity Commission's regulatory priorities evolve. Attorneys are advising HR teams to review accommodation policies and complaint-handling procedures now rather than after a charge is filed.
| Risk Area | What Triggered Attention | Practical HR Response |
|---|---|---|
| Diversity program design | Federal settlement over alleged discrimination against white employees | Ensure programs are open to all groups, with eligibility criteria reviewed by counsel |
| Religious accommodation | Attorneys anticipate a surge in claims | Audit accommodation requests, timelines, and documentation practices |
| AI-assisted decisions | Managers revisiting overlooked employees; candidates screening employers with AI | Validate tools, document criteria, and audit outcomes by demographic group |
| Complaint handling | Rising scrutiny across multiple claim types | Confirm investigations are prompt, neutral, and consistently documented |
Developing Leaders Without Replacing Human Judgment
Leadership development is where many of these threads converge. Workday and The Josh Bersin Company recently brought the Galileo AI agent into Workday's Sana platform, giving people leaders real-time guidance on hiring, team structure, and skill benchmarks, as described in a partnership announcement. Tools like this can put enterprise HR knowledge at a manager's fingertips.
Yet a separate workplace study cited in the roundup offers an important caution: AI-driven coaching tools are increasingly used to groom high-potential leaders, but they deliver their best results only when reinforced by human mentorship. Experts warn that replacing interpersonal feedback with software entirely can erode employee trust and engagement. That warning fits the promotion findings. A tool that prompts a manager to reconsider a candidate works best when the manager then has real conversations with that person about goals and development.
Budget decisions reinforce the point. Research from General Assembly, referenced in a roundup of L&D trends, shows that learning leaders are more than twice as likely to expect budget growth when their curricula align with documented business goals. As AI upskilling expands, boards are beginning to treat workforce education as a strategic lever rather than a discretionary cost.
A Practical Framework for Leadership Pipelines
- Pair AI recommendations with structured human review. Require a documented conversation before any promotion decision is finalized.
- Audit outcomes, not just tools. Track who is nominated, reconsidered, and promoted, broken down by relevant groups.
- Tie development programs to business priorities. Link AI-enabled coaching to specific competencies the organization needs in the next 12 to 24 months.
- Keep mentorship in the loop. Assign human sponsors to every high-potential track, especially for employees the algorithm flags.
Looking Ahead: Fairness as an Operating Discipline
The week's stories point toward a single conclusion. AI is changing how managers decide, how candidates evaluate employers, and how leaders are developed. At the same time, legal scrutiny of diversity programs, religious accommodation, and workplace practices is intensifying. Organizations that treat AI as a neutral fix will be exposed when its outputs are challenged. Organizations that treat it as a disciplined input, governed by clear criteria, regular audits, and strong mentorship, will be better positioned to show that their decisions are both effective and fair.
For U.S. HR leaders, the next step is to examine the full decision chain: how talent is evaluated, who reviews those evaluations, how programs are designed, and how complaints are handled. The companies that get this right will not just avoid legal trouble. They will build the kind of trust that keeps strong employees, and strong candidates, choosing to stay and to apply.