Governance, Ethics & Legal Practical insights
Track usage rights for AI campaign assets clearly
Build an asset-rights record for AI campaigns covering reference inputs, supplier terms, intended use and unresolved copyright questions before release.
An AI-generated file still needs an asset-rights review. The team should be able to explain what went into it, which terms governed generation and where the result will appear. NIST identifies intellectual-property issues as a generative AI risk area; its framework is not legal clearance for a particular campaign. [1]
Record the starting material
List reference photographs, logos, music, scripts and customer-supplied files. Attach the source and applicable permission or licence. Distinguish permission to display a photograph from permission to upload it to a generative service or use it to produce variants. Ask the rights specialist when the scope is uncertain.
Capture the terms that applied
Save the vendor, service tier, generation date and relevant contractual terms. Identify restrictions and any indemnity conditions rather than summarizing them as ‘commercial use allowed.’ Supplier permission to use an output does not settle every possible third-party claim, and copyright protection of the output itself is a separate question.
Review the intended placement
Check whether the asset will appear in paid advertising, product packaging, social posts or a customer case study, and in which territories. Inspect the generated result for recognizable brands, distinctive protected elements and unintended likenesses. A generic prompt does not guarantee a generic output. Escalate uncertainty with the actual image or audio attached, not just the prompt.
Keep rights with the file
Store the approved scope alongside the final asset in the media library. When an editor crops, localizes or reuses it, the rights record should remain easy to find. Set review dates where permissions expire. This process will not resolve every jurisdiction-specific copyright issue, but it prevents a common operational failure: losing the evidence and conditions needed to make an informed publication decision.
Sources and evidence
Sources checked on 4 October 2026. Proposed workflows and hypothetical examples are editorial analysis.
From insight to practice