Two governments. One week. The same direction.
The UK House of Lords Communications and Digital Committee published its report “AI, copyright and the creative industries” on March 6, 2026. Four days later, on March 10, the European Parliament adopted a resolution on copyright and generative AI. According to Slaughter and May’s legal analysis, both are strongly pro-copyright, favoring licensing and remuneration over broad exceptions that would allow AI developers to train on copyrighted content without payment.
The distinction in legal weight matters. The Lords report is a committee recommendation, not legislation. It carries no binding legal force. The EP resolution is a non-binding parliamentary position. Neither is enacted law. What they represent is political signal, and when two major jurisdictions signal in the same direction within a week, that signal carries weight for compliance planning.
The report addresses the core commercial tension in AI training data: should AI developers be required to license copyrighted content, or should broad text-and-data-mining exceptions allow them to use it freely? The Lords committee comes down firmly on the licensing side. The report also recommends, according to analysis of its contents, statutory transparency requirements for large AI developers and new protections against uses that replicate creative styles.
The UK Government was reportedly expected to publish an economic impact assessment by March 18, 2026, according to reports, though that publication had not been confirmed at time of writing.
For AI developers using copyrighted content in training pipelines, the compliance trajectory in both the UK and EU is moving toward mandatory licensing. That’s a different posture than the US approach, and this week made the divergence clearer.