AI update: Regulations, Government Safety Summit, and IP rights concerns

International AI regulations

The EU’s plans for an AI Act are considered to be ahead of the curve in terms of regulation. Talks on the final form of the law are in progress, with approval expected by the end of this year. A two-year grace period will allow companies and organisations to adapt, with the EU AI Act in full effect from 2025. Across the pond in the US, future legislation is anticipated at state, local and federal levels, following publication of a blueprint for an AI Bill of Rights and the White House is calling upon policymakers to “codify these measures into law”.  

Closer to home, however, the UK government mooted a light touch and ‘pro-innovation’ approach to regulation of AI. Its white paper released in March 2023 outlined five principles for regulation as: 1) safety, security and robustness, 2) transparency and explainability, 3) fairness, 4) accountability and governance, and 5) contestability and redress. Rather than passing formal legislation, however, the Government proposed to support regulators such as OFCOM to incorporate those principles in their work. Consistent with this approach, UK advertising regulator the ASA recently published guidance confirming that it would not introduce any new rules regarding AI, but noted that how an ad was created may be relevant in assessing compliance with its rules (for example, due to bias that generative AI may create).

 

Calls for protection against AI

Tech and industry experts fear, however, that UK policymakers have underestimated the scale of the AI challenge. In recent weeks, government committees and industry bodies alike have called for stricter policing and regulation to protect consumers and society and/or preserve human innovation and creative works.

In August, the government’s Science, Innovation and Technology Committee set out twelve significant challenges AI presents, and that AI governance must overcome in order to ensure public safety and confidence in AI. These included bias, privacy, misrepresentation, transparency, intellectual property and copyright, employment and even existential threat to human life. The committee’s chair, Greg Clark, believes that contrary to the Government's previous approach, legislation before the next election is ‘eminently feasible’.

Meanwhile, the Culture, Media and Sport Committee urged Government to heed the “chorus of warnings from musicians, authors, and artists” of failing to protect intellectual property in the emerging AI landscape. In its second report on connected technology, the Committee called on Government to abandon a proposed exemption that would give AI developers free rein to use copyrighted songs and books for training AI; a proposal which CEO of UK Music, for example, has described as a ‘green light to music laundering’.

The strength of feeling in creative industries regarding AI is undeniable: from Hollywood writers striking over concerns of safeguarding jobs, to the music industry criticism and publishing calls for controls. Early in September, 26 global publishing and journalism organisations united in “an unprecedented collaboration that safeguards the interests of content creators, publishers, and consumers alike” to release a set of Global Principles for AI. The Principles call for the ethical development and deployment of AI systems and applications, subject to accountability requirements and legislation that will protect publishers’ intellectual property, brands, consumer relationships, and investments.

 

Government agenda for AI Safety Summit

Ahead of the AI Safety Summit in early November, the Government has set out five ambitions which will frame the summit’s discussions. Built on initial stakeholder consultation and evidence-gathering, the agenda is: understanding the risks posed by AI, international collaboration for AI safety, safety measures at organisational-level, AI safety research and showcasing how safe development of AI will enable AI to be used for good globally.

While safety is understandably high on its agenda, how or when the Government may respond to the questions of intellectual property and recognising human creative endeavour is unclear. Given the policymakers’ original ‘innovation-focused’ programme, and pressure from the Culture, Media and Sport Committee, action to preserve of the value of human creativity and address fears that AI can undermine creator’s rights and compensation may perhaps be expected after the AI Safety Summit.

In any event, a greater level of regulation and legislation for AI than previously suggested now seems to be a likely outcome.

 

Protect your business IPs

Until parliamentary policy becomes clear, IP-generating businesses and consultants, particularly in the creative industries, will do well to maintain records of IP used and created, as well as clear evidence of ownership and licensing, to support any assertion of their rights which may be necessary in future. If operating internationally, it will also be important to consider how diverging rules between the UK, the EU and the US may benefit or affect your business in the future. If you are using generative AI in your business, navigating the risks of the tools will be key to leveraging the benefits the technology offers.

 

Do you need support with IP management and protection?

If you need advice on IP strategy, management, licensing, or protection, our commercial & technology team can assist. Helen Dobson is a commercial and technology lawyer with expertise spanning the digital, telecoms, transport, healthcare and retail sectors. Her expertise spans strategic partnerships, the exploitation, and management of intellectual property, as well as IP licensing and assignment. Contact us today, on [email protected].