New UK bill seeks to address AI copyright issues

The Automated Online Software (Access and Transparency) Bill is a private member’s bill presented to the UK parliament by an MP for the opposition Conservative party, Damian Hinds. Out-Law has asked the Department for Science, Innovation and Technology if the bill has the government’s backing – something which will be central to whether the bill progresses through the parliamentary process and becomes UK law.

The specific drafting of the bill has yet to be published on the parliament website, but details of the proposals have been published by trade body the News Media Association (NMA), which has backed the bill Hinds has developed.

Read more of our coverage on AI and copyright:

• The UK AI policy pivot signalled by new copyright reports
• Drop plans for AI-related copyright exception, UK ministers urged
• ‘Workable’ AI copyright solutions lacking amidst UK policy ‘reset’
• Getty Images v Stability AI: Getty’s copyright case against Stability AI fails
• AI copyright regime steers away from requiring licences in all cases

In a statement, the NMA said that the new law proposed, if implemented, “would require the operators of bots that systematically access and copy content from UK websites to declare their identity and the purpose of their activity”.

The NMA said there has been an observed increase in bot traffic and highlighted that the data scraping those bots engage in is impacting visitor numbers to publisher websites and the underlying commercial business models in the industry.

“In a world of exponentially growing bot traffic and harvesting of data and content to build and operate AI models, bot transparency is vital to the health of the online ecosystem,” the NMA said. “Some responsible operators already recognise and embrace this by identifying the bots they use and the intended purpose of their bot activity. Such transparency imposes little burden on bot operators. This lack of basic transparency leaves website owners unable to identify bots and block malicious traffic. It prevents them from negotiating a fair price for access to their valuable content and can even force their websites offline altogether.”

Hinds, a member of the House of Commons’ Culture, Media and Sport Committee, said bot activity poses a threat to journalism.

“My bill will do one simple thing: if you run an online bot that accesses a website and takes content and data, you have to say who you are and what you’ll do with what you take,” Hinds said. “This isn’t heavy-handed regulation, and it doesn’t seek to regulate AI models or dictate behaviour. It just requires basic transparency… If news outlets can’t secure fair remuneration for their work, through subscription or ad revenue, journalism will become unsustainable – and there’d be nothing left for the bots to scrape from.”

The bill represents the latest attempt by some UK lawmakers to add specific AI-related copyright protections to UK statute amidst concerns that AI developers are scraping content published online and using it to train their AI models.

Last year, some peers fought to add amendments to the government’s then Data (Use and Access) Bill to force the government to write new regulations to deliver greater transparency over use of copyright works by AI developers. The government successfully resisted those efforts, but its concessions resulted in a finalised Data (Use and Access) Act that imposed a statutory timetable on it to report on – and make recommendations in relation to – AI and copyright matters.

The government published its AI and copyright reports in March this year. However, it confirmed that it would not pursue any immediate changes to UK copyright law relating to AI training and development and resisted calls from lawmakers and industry to commit to a timetable for follow-up action on the work it did pledge to undertake. That work includes some further evidence-gathering over the impact of copyright law on AI development, assessing the impact of rules in place in other jurisdictions, supporting industry-led development of tools and standards – such as around control of access to and use of works – and keeping market-led licensing approaches under review.

In January, government ministers Liz Kendall and Lisa Nandy admitted before a parliamentary committee that “workable” solutions that enable transparency over the content and data used to train AI models and allow rightsholders to opt their works out from being used for that purpose, had yet to be found.

Cerys Wyn Davies of Pinsent Masons, who specialises in brand protection and creative rights in the AI age, said: “Visibility as to the extent of unauthorised scraping of content by online bots would be helpful in the analysis of the issue faced by both content providers and AI developers, and should also assist the government in its essential evidence gathering. However, the impact of such scraping can only be addressed if copyright law in the UK and in other jurisdictions is clarified and workable licensing arrangements endorsed.”

Intellectual property law expert Gill Dennis, also of Pinsent Masons, said: “The proposal reflects a growing focus in the UK on transparency as the linchpin of any workable AI-copyright framework. That aligns with wider policy discussions emphasising that rights holders need visibility over how their content is accessed and used before meaningful licensing markets can function.” 

“However, the effectiveness of any UK measure will ultimately turn on enforcement, particularly given the global and often opaque nature of automated scraping activity. Even with transparency obligations, identifying the true operator behind bots and pursuing remedies across jurisdictions is likely to remain complex and resource‑intensive. Without parallel progress on technical standards and cooperation mechanisms, there is a risk that compliance burdens fall on publishers while bad actors remain difficult to police in practice,” she added.

The second reading of the Automated Online Software (Access and Transparency) Bill is due to take place on Friday 16 October.