Special offer. See more information about Outbyte and uninstall instructions. Please review EULA and Privacy policy.

Disclose AI use truthfully and specifically: say what work it affected, what the tool did, and what you reviewed, changed, or validated. First check the rules for your employer, client, role, industry, jurisdiction, and deliverable; there is no universal requirement in the sources cited here to disclose every AI-assisted workplace task. Keep the record somewhere authorized colleagues can retrieve it, and do not enter sensitive or non-public information into an unapproved public AI tool.

Check the rules that apply before you disclose

Disclosure requirements depend on the work and context. Your employer or client may set a policy, and requirements may also come from a regulator, funder, publisher, or partner organization. Check the applicable rules before choosing whether, where, and how to disclose. The NIST AI RMF Govern Playbook notes that legal and regulatory requirements vary by application and context. The CDC guidance addresses scientific work and directs readers to follow relevant organizational and partner requirements; it is useful as a disclosure model, not a universal employment rule.

For EU-related work, be precise about what transparency rules cover. The European Commission says AI Act Article 50 transparency obligations apply from August 2, 2026. Its materials describe marking certain AI-generated content and labeling deepfakes and certain AI-generated text publications informing the public on matters of public interest, subject to conditions. They do not establish a blanket rule that employees must disclose every internal AI-assisted task. See the Commission’s Code of Practice page and guidelines for the relevant scope. The Code is a voluntary compliance tool; the underlying transparency requirements are legal obligations.

What to include in a useful disclosure

For substantive AI assistance, make the disclosure concrete enough that another person can understand what happened and what you did. CDC’s scientific-work guidance organizes disclosures around “Content Affected + Action Taken + AI Tool + Purpose of AI Use + Human Oversight.” Adapt those elements to the workplace setting without presenting them as a universal employment requirement.

Special offer. See more information about Outbyte and uninstall instructions. Please review EULA and Privacy policy.
  • Work affected: Identify the document, code, analysis, or specific sections or components.
  • Action and purpose: State what the AI did and why you used it, such as drafting, editing for clarity, summarizing, or helping explore an approach.
  • Tool: Name the platform and, if available, the model or version.
  • Your oversight: Describe what you checked, revised, tested, or validated. Do not claim checks you did not perform.
  • Responsibility: Make clear what you contributed and who is responsible for the final work under applicable policy.

The U.S. Department of Labor’s AI Literacy Framework says workers should use their expertise, context, and discretion when interpreting, using, or revising AI-generated content. It supports accurately describing your contribution, but does not prescribe a disclosure template.

Use a truthful, specific sentence

Adapt this example only to facts that are true and consistent with the rules for your work:

I used [tool and model/version, if known] to [action] on [specific work or sections] for [purpose]. I reviewed [what you checked, changed, or validated] and remain responsible for the final result.

For example, if an approved tool helped refine the wording of a report introduction, say that it assisted with wording in that section and describe the review you actually completed. Avoid vague statements such as “AI helped” when readers need to know what it affected, and avoid implying that you independently created material that the tool generated.

Special offer. See more information about Outbyte and uninstall instructions. Please review EULA and Privacy policy.

CDC’s sample for scientific manuscripts is narrower: it describes language refinement in specified manuscript sections and says the authors reviewed and approved the edits. That example is guidance for scientific work, not a required workplace formula. Read the CDC guidance for its context and sample wording.

Document the work in an approved, retrievable place

Keep a record that fits your organization’s process and can be accessed by authorized people who need it. NIST recommends that organizations establish documentation policies and appropriate storage and access procedures. Its playbook discusses organizational AI governance; it does not require every employee to maintain a personal log or paper notebook.

Depending on your organization’s rules, a record may capture the tool and model/version if known, the task and affected work, the purpose, and the review or validation you performed. For code, analysis, or other methodological work, preserve prompts, settings, inputs, and validation steps when needed to support reproducibility and permitted by security requirements. CDC’s advice on those details is specifically for scientific work.

Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Support on Ko-Fi

Protect sensitive and non-public information

Do not put sensitive, protected, or other non-public material into a public AI tool unless your organization has approved that use. CDC advises against entering such data into public AI tools, while NIST recommends connecting AI governance with data governance, especially for sensitive or higher-risk data. Check your organization’s approved-tool and data-handling rules before using AI on work material.

Special offer. See more information about Outbyte and uninstall instructions. Please review EULA and Privacy policy.

Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.