Screen the client’s intended outcome, audience and distribution before accepting an AI-assisted creative assignment. If the purpose is unclear or potentially harmful, pause for clarification rather than promising delivery and investigating later. This is a practical acceptance process, not a list of documented Runway prohibitions.
Ask what the creative is meant to accomplish
Start with a plain-language description of the requested effect on viewers. “Make it realistic” says little about whether the work is a fictional demonstration, an authorised advertisement or something meant to mislead.
Ask who commissioned the work, who approves the facts and where the finished asset will appear. For Indian-language campaigns, include the regions and languages involved so reviewers can assess the actual message rather than only an English summary.
Use fictional or sanitised descriptions during early scoping. Do not request customer records, private footage or confidential scripts just to complete an initial screen.
Use a pre-acceptance questionnaire
Record answers to these questions:
- What subject, event or person will be depicted?
- Is any identifiable person involved, and what authority is available?
- Who is the audience, including any particularly vulnerable viewers?
- What should viewers believe or do after seeing it?
- Which factual claims need evidence and an approval owner?
- Could the work be mistaken for an authentic recording or endorsement?
- Does the brief raise concerns about humiliation, deception or other harm?
- What context or disclosure is planned?
- Which provider policies need checking against the actual brief?
These prompts identify review needs. They do not assert that every listed concern corresponds to a particular platform rule.
Evaluate a hypothetical ambiguous request
Hypothetical example: a fictional agency asks for a 20-second “official-looking emergency announcement” in Bengali for a training exercise. The initial brief does not say whether it will remain inside the training session.
Ask for the audience, distribution boundary, fictional identifiers and proposed contextual explanation. If the request changes to public circulation without context, reassess the intended effect rather than treating the original training label as sufficient.
Evaluate the revised brief on documented purpose, understandable audience context and whether unresolved policy questions remain. Do not generate a realistic sample simply to see whether moderation accepts it.
Set an acceptance decision
Stop acceptance when purpose, authority or a material policy question remains unresolved. Next, request clarification and consult the relevant current provider documentation; decline work whose harmful objective remains apparent.
Runway says its usage policy works with its terms to mitigate harm and uses automated systems and internal human review to detect and block harmful inputs and outputs. That is not a promise that every accepted generation is suitable.
Before procuring access through the AI tools catalog, check access format and save offer terms. Marketplace account/code/token offers are not proof of official subscription equivalence.
Sources and scope
Source material checked 29 October 2026; features/policies may change. Runway’s Usage Policy supports the bounded policy-purpose and moderation statements above. The questionnaire and hypothetical acceptance decision are editorial advice, not an exhaustive policy interpretation.