72% of employees think AI is grading them. 73% of executives don't trust HR. And half of you are about to lose your best AI talent. Strap In, This is One Story.

I've spent the last week reading research that should have set off fire alarms in boardrooms. Most leaders are still busy naming their AI agents and rolling out dashboards nobody trusts.

Let me walk you through what's happening.

Kickresume surveyed 1,365 employees across six continents. 72% of them now suspect their manager is using AI to evaluate their performance. Almost one in five believe their review is mostly or entirely AI-written. The suspicion is loudest in Asia at 82%, with Europe and the U.S. right behind at 68% and 67%.

The same report found 71% of employees edit their self-reviews. They soften rough edges. They hide career goals. 36% avoid giving feedback about management because they don't trust the process to protect them. 24% would tell you, if it were anonymous, that they're overworked and under-recognized. 20% would say they're "barely hanging on."

This is what I call performance theater. The script gets read every quarter. Nobody believes the lines. Everybody plays their part anyway, while the employees update their resumes and look for the next job that pays them more.

Mercer's 2026 Global Talent Trends study surveyed 12,000 executives, HR leaders, investors, and employees. 57% of the C-suite ranks people analytics as the #1 ROI driver for 2026. Only 27% of those same executives believe their HR team can actually advise them on human capital risk.

HR Executive named the gap: insight theater. Same lane as performance theater, different vein. Pretty dashboards describing what already happened. Boardrooms asking what to do next. Crickets in between with a side of "OOOO Look ALL THE Pretty Colors."

By 2027, half of enterprises without a people-centric AI strategy will lose their top AI talent to competitors who actually invest in their workforce. Only 20% of leaders believe their people are truly AI-ready. Gartner calls the gap between "we rolled out an AI tool" and "our humans got better with AI" the enablement illusion.

Then there's BCG's randomized study of 1,261 managers. When companies framed an AI agent as a teammate (named it, stuck it on the org chart), accountability dropped 9 points. Error detection collapsed by 18%. Managers escalated work upward 44% more often because they stopped trusting their own judgment. Adoption didn't budge. So, we are seeing major decreases in productivity. Yay AI(*).

Read these four findings back to back, and the pattern is impossible to miss. (I might have mentioned a few times before that I am excellent at pattern recognition - it is one of the bipolar perks:))

Employees suspect their reviews are fake. The C-suite suspects their analytics are fake. The workforce suspects AI is replacing them rather than equipping them. And companies are pouring money into tools that make all three suspicions worse - the price tag is colossal.

The common thread is trust. The lack of it. The slow erosion of it. The expensive theater is being staged to cover for it.

The companies winning the next 18 months are the ones whose humans got measurably better with AI in their corner.

That requires three things most organizations don't have yet:

Real behavioral data, instead of survey data. Annual engagement surveys and pulse checks tell you what employees said six weeks ago. You need what's happening in the meetings your leaders run today.

Coaching that lands before the next call. Performance reviews are autopsies. Leaders need pre-flight checks.

Intelligence the C-suite actually uses. Trust scores. Behavior shifts over time. Risk tied to dollars. With this data, who needs 47 KPI tiles nobody opens?

This is exactly why we built Relate, and why Sandi (our AI meeting analyst) doesn't pretend to be a colleague. Sandi measures the four drivers of trust in every meeting your leaders run. She gives them coaching they can use before their next call. She hands CHROs the kind of human capital intelligence the C-suite has been asking for since Q1.

The opposite of theater is the data behind real, behavioral indicators that build real growth over time.

The next 18 months will sort companies into two groups.

The ones who treated AI like a headcount swap will spend 2027 explaining to their board why their best people left, and their analytics still don't predict anything.

The ones who used AI to make their humans the strongest version of themselves yet will spend 2027 quietly poaching everyone else's talent.

I know which side I'm on. Do you?