Copyright and AI: Consultation Results Signal Strong Mandate for Protections
Government's consultation on copyright and artificial intelligence has landed with a clear message: the creative industries want protections upheld and, if anything, strengthened. Equity has called the results a "vindication for the creative industries" and urged ministers to act on the consensus.
More than 11,500 responses were received. According to the progress report, a decisive majority backed licensing requirements for text and data mining (TDM) rather than creating new broad exemptions.
What the consultation tested
- Option 0: Do nothing (keep current law)
- Option 1: Strengthen copyright by requiring licences in all cases
- Option 2: Broad data mining exception
- Option 3: Data mining exception with a rights reservation and transparency measures (government's preferred option)
What respondents said
- 88% supported Option 1 (require licences in all cases)
- 7% supported Option 0 (no change)
- 3% supported Option 3 (exception with rights reservation)
- 0.5% supported Option 2 (broad exception)
- 1.5% did not indicate a preference
Equity highlights that 95% backed strengthening or maintaining protections. Its general secretary Paul W Fleming said the government must "listen to the weight of the unequivocal view of consultation respondents" and argued the preferred data mining exemption "must be dead in the water."
Why this matters for government officials
The signal from respondents is unambiguous: licensing should be the default for TDM that trains or powers AI systems. For policy teams, that points toward reinforcing current copyright frameworks rather than introducing a new, general-purpose exception.
For delivery bodies and contracting authorities, this has immediate operational implications for procurement, due diligence, and compliance when deploying AI tools or commissioning AI-enabled services.
Immediate actions for departments and agencies
- Procurement: Require vendors to evidence lawful data sources and model training practices (licences, permissions, provenance records).
- Contracting: Include warranties and indemnities on copyright compliance, plus audit rights on training data and embedded third-party tools.
- Risk management: Update AI risk registers to include copyright exposure, especially for generative and analytics tools that rely on TDM.
- Governance: Set clear guidance for staff use of AI systems, including rules for uploading internal or third-party content to external tools.
- Engagement: Continue dialogue with rights holders, unions, and industry bodies to align on practical licensing routes and transparency standards.
- Transparency: Ask suppliers for explainability on data inputs, model updates, and rights reservation mechanisms where applicable.
Context from the creative sector
Equity states there is clear consensus across the creative industries on the need for protections. The union also notes that members in film and TV are being balloted on whether they are prepared to refuse digital scanning on set to achieve adequate AI protections. The ballot closes at 12 noon on Thursday 18 December, with results expected around 1 pm the same day.
Fleming added that Equity will "fight for protections for their professional work through all means possible."
Policy takeaway
The consultation feedback points toward a licensing-first model for TDM. That aligns with a practical route for innovation that respects rights, clarifies responsibilities, and reduces legal uncertainty for government users and suppliers of AI.
Resources
Your membership also unlocks:
