LLM copyright and IP theft
A committee of British lawmakers has criticized the British government for what it sees as a passive approach to copyright infringement caused by large language models (LLMs).
In a letter addressed to Science and Technology Secretary Michelle Donelan, the House of Lords Select Committee on Communications and Digital criticized the government’s handling of copyright issues, claiming that its limited actions are effectively causing the government to commit copyright infringements approves.
The news comes just months after the country hosted a global AI summit where 28 countries signed the Bletchley Agreement to develop safe and responsible AI.
The UK government is not doing enough to protect itself from AI copyright issues
Since ChatGPT’s public preview launch in November 2022, and hundreds of subsequent launches of other similar generative AI tools, concerns have been raised about the use of copyrighted material, including text, images and audio data, in the development of the models .
The committee’s letter reads: “The government’s reluctance to take meaningful action amounts to a de facto endorsement of the practices of technology companies.”
The letter contrasts the government’s approach to copyright enforcement (“a failed series of roundtables”) with the high-profile attention and significant funding allocated to AI safety initiatives, such as the recently unveiled £400 million in funding for a new AI Safety Institute.
In an earlier report, the House of Commons Culture, Media and Sport Committee also said the UK government had failed to adequately protect the creative industries from copyright infringement.
This isn’t the first time generative AI has come under fire over copyright issues. Early in the collaboration between Microsoft and OpenAI, GitHub (owned by Microsoft and using OpenAI LLMs) was sued for lack of attribution.
In September 2023, almost a year after the public got their hands on generative AI at scale, Microsoft announced it would defend its Copilot users against copyright infringement lawsuits, indicating a high degree of certainty around responsible AI.
OpenAI followed in November 2023.
As the debate over copyright and AI continues, it is clear that a balance between technological innovation and the protection of IP rights deserves more attention.