Why ethics codes need more than a mention of AI

An AI mention is only the beginning of usable guidance

AI is absent from most organizations’ ethics codes, but adding a dedicated section can still leave employees unsure how to use it. Employees need guidance that connects decisions about the technology to their organization’s ethical commitments. So a code review needs to ask whether employees can use those commitments to decide what to do.

The same question matters when employees raise an ethics concern: a code needs to explain how the organization will respond. AI guidance and misconduct reporting involve different decisions, but both require employees to understand how a stated commitment works in practice. LRN recommends making those connections clearer; as an ethics and compliance services provider, it has a commercial interest in employers investing in such guidance.

Employees need clarity about AI readiness

For AI guidance to help employees make decisions, they first need to understand how their employer expects them to use the technology effectively. Recent HR industry analysis identifies insufficient employer guidance as one reason employees are uncertain about how to take advantage of AI. Guidance therefore needs to connect the organization’s plans with the choices employees face at work.

Employees and leaders can also have different expectations about those plans. In a report published in May, Adecco Group found that employees were less likely than leaders to expect their organizations to be ready to integrate agentic AI into workflows within a year. Agentic AI can take actions toward completing tasks, so bringing it into workflows affects how work gets done. Adecco Group, a workforce-services company, has a commercial stake in employers’ spending on workforce preparation and change.

Leaders can therefore examine two separate questions: how employees assess readiness and whether they understand what is expected of them. Leadership confidence in adoption plans leaves the second question open, so a code review needs to examine employee-facing guidance directly.

Explain what happens after an employee raises a concern

Clear expectations also matter in conduct reporting, where employees must decide whether and how to raise a concern. A reporting channel gives them a route to do so, while an explanation of the investigation process tells them what follows. This is separate from AI readiness, but it offers a concrete example of the practical information a code can supply.

The scale of reported misconduct makes that information important. An HR Acuity survey found that 55% of employees experienced or witnessed misconduct in 2025, compared with 41% in 2024, with the 2025 figure approaching a seven-year high. HR Acuity also found that 38% of respondents experienced multiple misconduct incidents. The annual comparison combines personal experience and witnessing, while the repeated-incident finding concerns personal experience; HR Acuity, an employee-relations technology provider, has a commercial interest in employers investing in misconduct management.

Alongside this increase in reported misconduct, low use of conduct codes gives employers a reason to examine how useful those codes are to employees. Relevant commitments have limited practical value if employees rarely consult them. Employers should therefore examine what prevents employees from using the guidance, while treating code use and misconduct as separate measures rather than assuming one caused the other.

Employees’ confidence in reporting adds another question to the review. LRN found that 66% felt able to report misconduct without retaliation, down from 71% last year. These figures measure perceived ability to report safely, so they help employers assess what employees expect when raising a concern. The frequency of retaliation is a separate outcome that requires its own evaluation.

Those expectations make it especially important to explain what follows a report. Most codes evaluated by LRN lacked explanations of their organizations’ investigation processes, leaving employees with a reporting route but little guidance about the next steps. LRN says employees “need more than a reporting channel” to surface misconduct or ethics concerns, and recommends explaining investigation processes in ethics codes. Employers can check for that omission directly, alongside employees’ confidence in using the channel.

Connect AI decisions to the principles already in the code

The investigation example shows why a commitment needs an explanation of how it works; LRN recommends applying that approach to AI guidance. In its assessment, most employers are unlikely to need entirely new conduct-code sections for AI. Instead, they may need to connect AI use to established principles, including accountability, fairness, transparency and sound judgment.

Before employees can use those principles in decisions, they need to understand what they mean in practice. An employee may understand that fairness matters while remaining unsure what it requires when using AI. Revised guidance should make the relationship clear: which principle applies, what judgment the employee needs to make and how the organization expects the principle to guide that judgment.

LRN describes this priority in its report: “The strongest codes will not simply add more content,” followed by, “They will help employees navigate emerging risks with practical guidance that is easy to find, easy to understand, and easy to apply.” These are LRN’s recommendations, reflecting its commercial interest in ethics and compliance guidance. They also give employers three distinct tests for reviewing a code.

The first test is whether employees can find the relevant guidance. Connecting AI to existing principles means making that connection visible to someone looking for help with the technology. A principle buried elsewhere in the document may be relevant yet hard to recognize as useful, so the review should examine how an employee would locate it.

Once employees find the guidance, the next tests are whether they understand it and can apply it. The explanation should connect the named principle to the decision at hand and give practical direction on emerging risks. Understanding the wording and using it to make a judgment are separate tasks, so reviewing both can reveal where a broad commitment needs a clearer explanation.

Design AI policy to help employees work, and revise it as circumstances change

Guidance that employees can apply also needs to explain what AI is intended to help them achieve. Sources who previously spoke to HR Dive advised organizations to center AI ethics policies on helping employees do their best work. This gives the policy an enabling purpose alongside its protective role: employees need to understand the intended benefit and the ethical expectations attached to using the technology.

The same sources advised organizations to address how they will prevent AI from eroding trust. So an explanation of employee benefit needs to sit alongside an account of how the organization will protect confidence in AI’s use. Making both responsibilities explicit helps employees understand how the organization intends to support their work and address risks to trust.

Those explanations need to remain useful as circumstances change. The sources told HR Dive that HR should treat AI ethics policies as living documents, revising them over time in the same way as broader conduct codes. HR’s responsibility therefore continues after publication: as the organization’s use of AI and its expectations change, the policy needs to reflect the decisions employees now face.

Key executive takeaways

• Test whether guidance is usable: Ethics and compliance owners need to assess whether employees can find, understand and apply code guidance to workplace decisions. A reference to AI has practical value when employees understand how it affects their work.
• Check employee readiness directly: Adecco Group found employees less confident than leaders about readiness to integrate agentic AI. Those responsible for adoption need to assess employee expectations separately from leadership confidence and clarify how employees are expected to use the technology.
• Explain what follows a misconduct report: Most codes evaluated by LRN lacked explanations of investigation processes. Ethics and compliance owners can address this gap while separately assessing reporting confidence, which fell from 71% to 66% in LRN’s findings.
• Connect AI decisions to established principles: LRN recommends linking AI use to accountability, fairness, transparency and sound judgment. Code revisions need to explain which principle applies to a decision and how employees are expected to use it.
• Keep AI policies current: HR needs to revise AI ethics policies as organizational uses and employee expectations change. The guidance needs to explain how AI supports employees’ work and how the organization intends to protect trust.