UK government releases AI action plan

The UK government has released its “AI Opportunities Action Plan”, outlining 50 ways it will build the AI ​​sector and turn the country into a “world leader”. The strategy includes expanding public computing capacity twenty-fold, creating a training data library, and building AI hubs in non-industrial areas.

Innovation is front and center in this new plan, marking a clear shift from the risk-averse approach of the previous Conservative government, exemplified by its AI safety summits and safety pledges. Most of the recommendations focus on developing AI infrastructure, promoting adoption, enhancing talent, and attracting investment.

Prime Minister Keir Starmer said, “Our plan will make Britain a world leader.” Press releaseThe emphasis was on creating “more jobs and investment in the UK, more money in people’s pockets and transformed public services”.

Scheme formed to promote use of AI across the country to ‘win the global race’

There are strong arguments in favor of this opportunity for the technology sector. In August 2024, the number of tech startups founded in the UK saw a “significant decline” for the first time since 2022. This metric is seen as an indicator of the industry’s growth – or lack thereof.

According to research by Stanford University, the UK ranks third in the world in terms of AI readiness, far behind the US and China. Tech giants like Google have also spoken out about laws in the UK that prevent AI models from being trained on copyrighted material and called for a “pro-innovation regulatory framework” to prevent the country from being left behind.

WATCH: UK government announces AI projects worth £32 million

On the other hand, evidence suggests that there is still significant room for improvement in AI safety and regulation. A report by Microsoft found that almost half of UK SMEs do not use AI technologies in any capacity, and 72% cited concerns about their potential unreliability as a barrier to adoption. In October 2023, research from the University of Cambridge ruled that UK needs AI law for safety and transparency So that companies can confidently invest resources in AI development.

Government entrusted the work to technical entrepreneurs and Newly appointed AI opportunities advisor to matt clifford develop action plan In July, which he Discussed with venture capital firmsTheir 50 recommendations on how to develop AI and promote its adoption will be implemented into a UK plan.

According to the International Monetary Fund, the AI ​​Action Plan could increase annual productivity by 1.5% and boost the economy by an average of £47 billion over a decade. Furthermore, Microsoft research found that adding just five years to the time it takes to implement AI in the UK could reduce its economic impact by more than £150 billion in 2035.

The Prime Minister said: “The AI ​​industry needs a government that is on their side, that will not sit back and let opportunities slip through their fingers. And in the world of fierce competition, we cannot survive. We must move fast and take action to win the global race.

Clifford’s key recommendations

Clifford’s proposals fall under three broad categories: laying the foundations for AI to flourish, boosting AI adoption across the public and private sectors, and keeping the UK ahead. Thirty recommendations belong to the first category, including:

• Establish “AI Growth Zones” in non-industrialized areas: Within these regions, planning requests for data centers will be expedited, and AI infrastructure will have better access to the energy grid, ideally from clean sources such as nuclear fusion. This is needed as the construction of new data centers in the UK has been halted due to inadequate power supply. Three private tech companies have already pledged £14 billion.
• Increase public compute capacity twentyfold by 2030: Clifford found that this would give the UK the processing power needed to fully adopt AI. By November 2022, the UK had only 1.3% of global computing capacityWhereas Microsoft ranked the country 11th in the world for cloud infrastructure in May. The initiative will be launched by the construction of a new supercomputer, a change since the government spent £1.3 billion building these resources in August.
• Create a National Data Library: This will involve collecting “five high-impact public datasets” to be made available to private AI researchers, but there is little clarity on how this will be achieved. “Responsibly, safely and ethically,” As claimed. Clifford also recommends creating a “copyright-free British media asset training data set”, which could be licensed internationally. This is unlikely to be accepted by the creative industries, which just last month called for greater protection of copyright laws so that artists retain control when licensing to AI firms.
• Be more aggressive with text and data mining: Similarly, Clifford says that “the current uncertainty around intellectual property is hindering innovation and undermining our broader ambitions for AI.” He recommends improvements in text and data mining practices. While he mentioned giving up control over the use of their content to rights holders, the mandate suggests this is not a priority. The government has started consultation on this recommendation.
• Regulators need to declare how they support AI innovation: Data regulators are too risk-averse, from Clifford’s perspective. They believe they should take active steps to support the development of AI, such as granting more licenses and AI resources and reporting to them annually. If reporting mandates and deadlines do not provide enough pressure, he suggests appointing a new central body with a “higher risk tolerance” to make such decisions.
• Nurture AI Talent: The AI ​​Action Plan includes a number of recommendations to support AI talent in the UK, including assessing the skills gap, supporting higher education institutions to teach relevant skills and promote AI graduates, expanding the number of AI education pathways This includes using the immigration system to attract graduates. International universities, and actively promoting diversity. In fact, only 28% of Coursera’s Generative AI course enrollments Are from women.

SEE: Red Hat: AI is the most in-demand skill in the UK for 2024.

Details about the three categories of recommendations

Compared to strategies that promote innovation, this first category of recommendations has relatively little nod to AI safety. These include continuing to support the AI ​​Security Institute and its research and building assurance tools.

The second category of the AI ​​Opportunities Action Plan outlines how, once the foundations have been laid, the government can promote the adoption of AI across the public and private sectors. Clifford recommends adopting a “scan → pilot → scale” approach whereby the government identifies high-impact opportunities for AI in the public sector, prototypes a solution, and then scales them across industries and sectors . It also suggests making digital government infrastructure available to private tech companies, creating an AI knowledge hub with guidance for both the private and public sectors, and addressing barriers to private sector user adoption .

The third category looks at how the government can keep the UK “at the forefront internationally” when it comes to growing the AI ​​industry. The category includes only one recommendation: creating a sovereign AI unit to support private sector AI research. It will provide companies with access to financial investment, calculations, data sets, foreign talent and the national security community “to maximize the UK’s potential to become a globally competitive national champion.”

Tech giants applaud UK’s AI approach, but fears remain of cyber attacks and copyright exploitation

Tech companies have generally responded positively to the AI ​​Opportunities Action Plan. They are reminiscent of British inventions of the past such as the locomotive and the Colossus computer to exemplify cutting-edge technology. They also cite the UK’s attitude towards AI as the reason for setting up shop here.

Alison Kay, VP overseeing UK and Ireland at Amazon Web Services, said the benefits that AI can bring and enhance are why it Pledged to invest £8 billion in data centers In the country.

Zahra Bahrolloumi, CEO of Salesforce UK & Ireland, told TechRepublic in an email that the UK is “prime position to fully unlock the opportunities of AI,” which is why it chose it as the site of its first AI center.

Naturally, not everyone agrees. Michael Adjei, director of systems engineering at data center security company Illumio, said researchers with access to the proposed national data library “will become prime targets for cyber attacks.”

WATCH: 87% of UK businesses are unprepared for cyber attacks

“Cybercriminals will seek to exploit hidden layers of AI, which are often proprietary and inadequately scrutinized,” he told TechRepublic in an email. “Vulnerabilities and coding flaws in these layers may remain undiscovered longer than in other AI layers, leaving them open to exploitation by third parties.”

The creative industries are also not sold on the new plan. Don Alford, executive director of the Society of Editors, said in a statement: “While we support efforts to accelerate development, the government must also support creative industries that continue to be exploited by unauthorized scraping of their content by generative AI tools.”

“The UK can achieve public service reform and seize all the growth opportunities associated with AI without facilitating US tech-led piracy of UK copyright works,” Dan Conway, CEO of the Publishers Association, said in a statement. statement,