In a pioneering move, United Kingdom Prime Minister Rishi Sunak has unveiled plans to establish the world’s first AI Safety Institute. This groundbreaking initiative will be dedicated to the evaluation, examination, and testing of various artificial intelligence (AI) technologies. The announcement was made during a speech delivered by Sunak at The Royal Society, in which he emphasized the need for a global effort to comprehend and address the potential risks associated with AI, while harnessing its benefits and opportunities for future generations.
The announcement comes just days before the UK is set to host the Global AI Safety Summit at the iconic Bletchley Park, known as the birthplace of computer science. Earlier this year, Sunak introduced plans for the UK government’s Frontier AI task force, with the objective of leading the secure and dependable development of cutting-edge AI models, including generative AI large language models (LLMs) like ChatGPT and Google Bard. Launched in June, the task force is supported by a substantial £100 million in funding, aimed at ensuring the nation’s sovereignty in AI capabilities and fostering widespread adoption of secure and reliable foundation models, thereby establishing the UK as a science and technology powerhouse by 2030.
In August, AI was officially recognized as a national security threat for the first time in the UK, following the publication of the National Risk Register (NRR) for 2023.
Sunak stated, “The British people should have peace of mind that we’re developing the most advanced protections for AI of any country in the world. I will always be honest with you about the risks, and you can trust me to make the right long-term decisions.” The AI Safety Institute is set to evaluate and research these risks, ranging from issues such as bias and misinformation to more severe potential risks, thus ensuring a comprehensive understanding of the capabilities of each new AI model.
Sunak emphasized the need for a shared understanding of AI risks, stating, “Right now, we don’t have a shared understanding of the risks that we face. Without that, we cannot hope to work together to address them.” To address this gap, the UK will propose the establishment of a global expert panel to publish a “State of AI Science” report. Sunak also underscored the importance of cooperation with AI companies, given their unique relationship with the UK, which has granted them privileged access to their AI models.
UK AI tsar Ian Hogarth previously cautioned that AI could be used by cybercriminals to launch attacks on the National Health System (NHS). This raised concerns about the potential risks of AI systems being utilized to launch cyberattacks on the health service or to design pathogens and toxins, as AI technology advancements make it easier for malicious actors to conduct attacks.
Hogarth stated, “The government is quite rightly putting these threats to the very top of the agenda, but technology leaders need to heed the warning and get moving, to better prepare for the next inevitable attack.” He also highlighted the international nature of the risks posed by advanced AI technology, emphasizing the need for global collaboration to tackle these global challenges.
In a world where AI is evolving rapidly, the UK’s commitment to enhancing the safety and understanding of AI is a significant step in safeguarding the future of this transformative technology.