AI completely absent from most organizations’ ethics codes, survey finds

What happened

A new survey from ethics and compliance firm LRN finds that most organisations have yet to update their codes of conduct to address artificial intelligence, even as AI tools become embedded in day-to-day work. The research, reported by HR Dive and CIO Dive, also shows a decline in the share of employees who say they actively reference or rely on their company's code of conduct, at a time when reports of workplace misconduct are rising.

Taken together, the findings point to a widening gap between how fast AI is being adopted inside organisations and how slowly governance frameworks — including the ethics codes meant to guide employee behaviour — are catching up.

Why it matters

Codes of conduct are one of the primary tools organisations use to set behavioural expectations and reduce risk, from data handling to decision-making. If AI use is largely absent from these documents, employees are left without clear guidance on acceptable use, disclosure, oversight or accountability for AI-assisted decisions — precisely as generative and agentic AI tools expand what employees can do with less human review.

The parallel decline in employees actively using their codes as a resource compounds the risk. A code that isn't consulted can't shape behaviour, regardless of what it contains. For leaders driving AI adoption or broader digital transformation, this signals that technology rollout is outpacing the policy, training and cultural reinforcement needed to make that adoption safe and trusted — both for customers affected by AI-driven decisions and for employees expected to use these tools responsibly.

By the numbers

• Most organisations' codes of conduct do not address AI, according to LRN's survey.
• Fewer employees report actively citing or using their code of conduct as a resource compared with prior findings.
• Rising reports of workplace misconduct were recorded alongside the decline in code usage, per the survey.

The Renascence take

It's tempting to read this as a compliance footnote, but it's really a service-design and trust problem hiding in plain sight. Codes of conduct are, in effect, an internal experience artefact — they shape how employees behave toward customers, data and each other. When that artefact goes stale, the gap doesn't stay internal for long.

Most organisations are treating AI governance as a policy-writing exercise rather than a behavioural one — and that's the mistake. A code nobody reads was never going to prevent misuse; the real lever is designing the moments where employees actually face an AI-related judgement call, and building reflexive guidance into those moments. Leaders serious about responsible AI adoption should audit where employees interact with AI tools today, then redesign guidance, nudges and escalation paths around those specific touchpoints — not just update a PDF nobody opens.