EXTRA One of the questions hanging over calls to require generative AI developers to obtain licenses for copyrighted works used in training their models, such as we saw last week from the Human Artistry Campaign coalition of rights owners, is whether legal foundation for that demand exists in current copyright law, or would require new action by legislatures to create that foundation. It’s a political question as much as a legal one. And while its legal colorings may favor the pro-licensing position, the political ones may not. Just how much they may not can be glimpsed in a new U.K. government report issued this month that illustrates how difficult it could be to persuade policymakers to provide any such new statutory authority.
The report, prepared by the office of the Chancellor of the Exchequer, covers a range of issues involving digital technologies with an eye toward developing a “pro-innovation” regulatory framework to promote Great Britain’s technological competitiveness and leadership. And the section on generative AI very much frames excessive deference to intellectual property rights as an impediment to that goal.
If the government’s aim is to promote an innovative AI industry in the UK, it should enable mining of available data, text, and images (the input) and utilise existing protections of copyright and IP law on the output of AI. There is an urgent need to prioritise practical solutions to the barriers faced by AI firms in accessing copyright and database materials. The government should work with the AI and creative industries to develop ways to enable TDM for any purpose, and to include the use of publicly available content including that covered by intellectual property as an input to TDM (including databases). The opportunity here is to focus on clarifying a simple process concerning the input to AI models; IP rights and their enforcement would apply to the output of any product. We also recommend a code of practice and a requirement for altered images to be labelled as generated or assisted by AI…
The government should recognise that the advent of generative AI globally represents both an opportunity and a challenge to the creative industries and education sectors but that the restriction of data access for training sets would be likely to put the UK at a disadvantage and impede domestic development of the technology [emphasis added].
Far from providing new statutory authority to support licensing, in other words, the report recommends the government move proactively to enable unfettered, unlicensed use of copyrighted works in training generative AI models.
The report’s recommendations are just that, of course — recommendations in a single report from a single government ministry. But as noted here in a previous post, if the question of AI licensing gets framed politically as one of copyright vs. innovation and technological leadership, rights owners could find themselves on the defensive with lawmakers eager to be seen as pro-innovation.
The report also underscores how high the stakes are in the lawsuits currently being prosecuted by artists, illustrators and photographers against Midjourney and other developers of AI image generators over the use of their copyrighted works in training the models. Early court victories, resting on long-established precedents, could help head off contrary moves by public policymakers. Early defeats in those cases, on the other hand, could spell trouble for rights owners.