Open source advocates in the UK have called on the government to ensure AI technologies have sufficient access to public data to support the growth of the UK AI sector.
In a joint letter signed by OpenUK, IP Federation, Wikimedia UK and the Entrepreneurs Network, among others, the government was urged to “avoid frictions around using data necessary to develop safe AI” by clarifying the intellectual property (IP) laws surrounding the technology.
According to the organisations, open access to public data is vital to ensuring AI safety goals established in the AI Safety Summit’s Bletchley Declaration are met.
The letter said that while “many other countries have clarified their intellectual property laws to support AI and innovation,” the UK has yet to explicitly support the knowledge transfer of AI.
Large AI models, including OpenAI’s ChatGPT, require access to huge volumes of data for training.
The letter has argued that while not all data should be considered usable for AI training, new models must have protected and open access to data that is lawfully available to the public.
“Any restriction on the use of such data or disproportionate legal requirements will negatively impact on the development of AI,” the letter said.
“Not only inhibiting the development of large-scale AI in the UK but exacerbating further pre-existing issues caused by unequal access to data.”
In support of this, the organisations have outlined a proposal for a code of practice to inform AI IP legislation.
The code of practice called for clarification on access to broad data sets that are publicly available, the recognition that other nations have their own policies which must be factored in and the avoidance of friction around access to necessary data for safe AI development.
Amanda Brock, CEO of OpenUK, said: “The open letter is an important call to action to make the UK’s IP System ‘AI-friendly’ for all AI. For our open technology community, this fair and transparent sharing is the lifeblood of our industry.”