Most leadership teams still believe company culture gets shaped in town halls, values workshops, and onboarding decks. That belief no longer holds. Deloitte's 2026 Global Human Capital Trends report found 42% of workers say their organisation rarely evaluates the impact of AI on people at all, even as AI quietly reshapes what "good work" means inside their teams every single day.

Deloitte has a name for what happens next: culture debt — the negative consequences organisations accumulate when they scale AI without redesigning leadership behaviour, accountability, and workplace norms. The numbers behind it are stark. 80% of leaders, managers, and workers in Deloitte's 2026 survey say they are concerned their colleagues are using AI to appear more productive than they actually are. That is not a technology problem. That is a trust problem, and trust is the actual foundation company culture is built on.

Is AI actually changing company culture?

Gallup's Q1 2026 workforce study found AI adoption is pushing organisational cultures in roughly equal positive and negative directions. In workplaces that have implemented AI, 25% of employees say their culture has worsened over the past year, while 24% say it has improved. Only 51% say it has stayed the same — notably lower than the 59% figure among employees in workplaces that have not adopted AI at all.

This is not a marginal shift. It means one in four employees at AI-adopting companies is watching their culture actively deteriorate, often without leadership fully understanding why. Deloitte puts the mechanism plainly: workers are quietly adopting or challenging new norms around effort, ownership, fairness, and accountability, because organisations are not answering the basic questions AI has raised. Is it cheating to use AI for a piece of work? What counts as effort now? Who is responsible if the AI gets it wrong?

How are AI prompts reshaping norms?

Company culture has always formed from small, repeated behaviours, not formal policy documents. Deloitte's own framing captures this well: culture is built through the meetings people cancel, the conversations leaders lean into, and how credit gets distributed. AI is now embedded directly inside those daily behaviours.

Several specific shifts are already visible in workplace data. Trust between coworkers is eroding quietly — 77% of workers say they review a colleague's work more carefully once they know AI was involved, and 36% say they review it "much more carefully." Rework is becoming routine, with 45% of workers reporting they have had to fix or redo a colleague's output because it relied too heavily on AI, and 57% of managers reporting the same problem with their own teams. AI use is also unevenly self-directed, skewing toward individual and team-level use rather than organisation-wide rollout — meaning norms are forming inconsistently, team by team, without a shared standard. And nearly a third of employees now sit in "survival mode," alongside a widening "capability chasm" where leadership signs enterprise AI contracts while the average employee quietly experiences tech anxiety.

---

Who is shaping this version of culture?

Gallup's research found manager support is the single strongest lever available. Employees who strongly agree their manager actively supports the team's AI use are far more likely to report their culture has improved — 31% versus 21% for those without that strong manager support. They are also three times more likely to say culture has improved "a lot."

This puts real weight on a group most companies have not adequately prepared. 57% of CHROs in Gallup's roundtable survey say they are now providing AI training specifically for managers — a clear signal that the gap is being recognised, even if it is not yet closed. Microsoft's 2026 Work Trend Index, surveying 20,000 workers across ten global markets, found organisational factors — culture, manager support, and talent practices — account for more than twice the impact on AI outcomes compared to individual mindset and behaviour. In plain terms: how a manager talks about AI shapes an employee's experience of it far more than the employee's own attitude does.

Does this global data apply globally?

The 2026 Work Trend Index deliberately sampled ten markets — including India, Brazil, Japan, and several across Europe — specifically to avoid a US-centric read on this shift. The finding that organisational culture and manager support outweigh individual AI skill held consistently across that global sample, not just in Western markets. This matters directly for CHROs running operations across Africa, the Middle East, and Southeast Asia, where AI investment is rising just as fast, if not faster, than in markets assumed to be leading.

The uneven, self-directed nature of AI adoption is unlikely to look different in fast-growing markets. If anything, regions where formal AI governance and manager training programmes are still being built face a higher risk of the same culture debt forming without anyone noticing — simply because fewer companies have the internal infrastructure yet to track it.

Conclusion

The most significant finding in this data is that company culture was never really under direct executive control in the first place. It has always been an emergent outcome of thousands of small daily behaviours. What AI has done is inject a powerful new behaviour into that system — how people use, trust, and talk about AI-assisted work — without most companies deciding in advance what they wanted that behaviour to look like.

The organisations managing this well are not the ones with the most sophisticated AI tools. They are the ones treating culture debt as seriously as financial debt: tracking it, naming it, and assigning someone to own it before it compounds. The real question is not whether AI is shaping company culture — that is already settled. The real question is whether anyone in the organisation is actually watching it happen, or whether the norms are simply being written, prompt by prompt, by whoever happens to be using AI the loudest.