For the first time, the UK Government has published a new report on the capabilities and risks posed by frontier AI.
Frontier AI is defined by the government as highly capable general-purpose AI models that can perform a wide variety of tasks to a similar or better standard than today’s most advanced models.
The report is split into three parts with a discussion on the need for further research into AI risk coming first. It covers frontier AI’s current capabilities, how they could improve, and their associated risks.
Second is a report on the safety of generative AI, which centres around how development in this area could bring great benefit – but it could also increase the effectiveness of attacks.
Last, is a report from the Government Office for Science that looks at the key uncertainties in frontier AI development, the risks related to future AI systems as well as a range of scenarios until 2030.
The paper was published in conjunction with the prime minister, Rishi Sunak, delivering a speech on the risks and opportunities of AI.
LIVE: My speech on the risks and opportunities of AI https://t.co/LfOlJvISvO— Rishi Sunak (@RishiSunak) October 26, 2023
The publication comes ahead of the world’s first AI safety summit next week at Bletchley Park, which will use the government’s paper as a talking point.
The summit will focus on the risks of AI, especially by “non-state actors” to perform cyber-attacks or design bioweapons. The idea that governments could lose control of AI, enabling it to act autonomously will also be discussed.
Michelle Donelan, the technology secretary, said: “This marks a watershed moment, as the UK becomes the first country in the world to formally summarise the risks presented by this powerful technology.
“There is no question that AI can and will transform the world for the better, from making everyday tasks easier, to improving healthcare and tackling global challenges like world hunger and climate change. But we cannot harness its benefits without also tackling the risks.”
The summit will also investigate the dangers posed by the integration of AI into wider society – i.e., the effect it could have on elections, crime and online safety.
Donelan concluded: “No country can do this alone, which is why we will be welcoming governments, academics, civil society groups and businesses to Bletchley Park next month to build a shared understanding of the risks while discussing how we can develop and use AI safely and responsibly so that it changes lives for the better.”
To access the full paper, click here.
Image credit: iStock