Your AI Policy Isn’t the Same as AI Governance

Subscribe Now

For the past two years, DEI and immigration were what employers spent the most time worrying about. This year, AI has replaced both, and it isn’t close.

Littler’s 14th Annual Employer Survey, pulling from more than 300 C-suite executives, in-house lawyers, and HR professionals found that 84% of business leaders now expect AI to drive changes to workplace policy and regulation. Last year, 42% said the same. DEI concerns dropped to 38%. Immigration fell from 75% to 49%.

AI didn’t just move up the list. It pulled away from everything else.

What shifted isn’t awareness. Most employers have known for a while that AI matters. What shifted is that they’re realizing they’re already using it across HR, recruiting, operations, marketing, legal and the question is no longer whether to adopt it. It’s whether they’re actually managing what they’ve already brought in. Only 6% of surveyed employers said they don’t use AI for any function. The rest are using it in ways that are often well ahead of any real process for overseeing it.

The 68% Problem

Sixty-eight percent of employers now have a formal AI policy in place. That’s a start, but it’s also the easier part of the problem.

A written policy can tell employees not to paste confidential data into a chatbot. It can set some basic guardrails. What it can’t do is tell you whether the recruiting tool you bought is quietly screening out applicants in ways that expose you to discrimination claims. It won’t catch the moment a performance system starts producing results that consistently disadvantage certain groups. It doesn’t answer who’s responsible when an AI-driven recommendation turns out to be wrong.

Policy tells you what’s allowed. Governance is about what actually happens; who makes the decisions, who checks them, and who answers when something goes wrong. Most organizations right now have the first and are treating it as though it’s the second. It isn’t.

Where the Risk Is Building

The places where employers are using AI most heavily are also where the decisions carry the most weight. Hiring, screening, performance reviews, productivity tracking; these directly affect people’s jobs and livelihoods. And the legal responsibility for those decisions stays with the employer, not the software vendor. When a third-party tool produces a biased outcome or a decision that can’t be explained, the employer is the one in front of the regulator or the plaintiff’s lawyer. Not the company that sold them the tool.

That’s not theoretical. The lawsuits working through the system right now around hiring tools, disability accommodations, and data privacy are going to shape what responsible AI use looks like in practice before any comprehensive legislation does. Employers waiting for federal regulatory clarity before tightening up their practices are likely to find the courts got there first.

State-level requirements are already in effect. New York requires disclosure when AI played a role in layoffs. Illinois mandates disclosure when AI is used in hiring or promotion decisions. California has its own requirements. If you operate across states, you may have obligations you haven’t mapped yet.

The Quieter Story

Only 15% of employers said they had cut or were planning to cut headcount specifically because of AI, lower than a lot of the coverage suggests. But the bigger shift is less obvious. Roles are being redesigned. Performance expectations are moving. What counts as a normal day’s output looks different than it did two years ago. Those changes still require real HR decisions around retraining, compensation, and job structure, and they carry fairness and legal implications of their own.

There’s also something worth noting about managers specifically. Fifty-five percent said they’re worried AI tools will lower their own pay this year, up from 44% last year. The people running your teams are anxious about the same technology their organizations are speeding up adoption of. That tension doesn’t go away on its own.

The Pattern

This is the same gap we see across workforce decisions more broadly. Organizations move fast on adoption and build the oversight structure later, usually when something forces it.

The organizations we see managing AI well right now aren’t the fastest adopters. They’re the ones who treated accountability as something to build alongside the technology, not after it. They know what tools they’re using, who owns the decisions those tools influence, how their vendors handle model updates, and what their answer is if they’re ever asked to explain an outcome.

Most organizations can’t do that yet. The Littler data makes the scale of the gap pretty clear. Sixty-eight percent have a policy. Far fewer have anything they’d comfortably call governance. The risk is sitting in that space building quietly, the way it usually does, until something brings it into the open.

If you’d like to talk through where your organization stands on AI governance and what your current practices may need to address, our team is ready to help.

Talk to the Converge Team

Sources: