JUNG YEON-JE/AFP/Getty Images
show image

Oscar Williams

News editor

DeepMind’s Demis Hassabis is set to advise the government’s new Office for AI

Demis Hassabis, the CEO of Google DeepMind, has been named as the first adviser to the government’s new Office for Artificial Intelligence.

In his new position, Hassabis will advise the government on building the skills and capability needed for the UK to embrace the benefits of the technology, while also minimising its risks.

He founded DeepMind with Shane Legg and Mustafa Suleyman in 2010 before selling it to Google for a reported £400m four years later. Based in London’s Kings Cross, the company is regarded as one of the world’s leading centres of AI research.

“I’m honoured to be taking on the role of Adviser to the Office for AI, and look forward to the huge opportunity that lies ahead,” said Hassabis. “I’ve always believed that AI could be one of the most important and widely beneficial breakthroughs of the 21st century – and as a proud Londoner, it’s fantastic to see the UK’s world-class universities and start-ups already making major scientific advances.

“Alongside the research, I’m very excited about the role the UK can play in making the case globally for AI’s safe and ethical deployment.”

Speaking at the opening of the London Office for Rapid Cybersecurity Advancement this afternoon, digital secretary Matt Hancock also announced that Dame Wendy Hall, an AI expert who co-authored a government report on the sector last year, and Tabitha Goldstaub, co-founder of CognitionX, will serve as AI skills and business champions respectively and oversee the government’s AI “grand challenge”.

“We want to harness the best possible AI leadership to help us seize this opportunity,” said Hancock. “Demis Hassabis, Tabitha Goldstaub, and Wendy Hall have the expertise and vision to help us make sure the huge benefits of this powerful new technology are available to everyone.”

Goldstaub, who recently authored a report into the state of AI in London, will also serve as the chair of the new AI Council. The industry forum will be charged with growing the UK’s ecosystem of AI startups and scaleups and promoting adoption of the technology across the private sector.

Artificial intelligence has the potential to transform our world for the better but to be successful we need to galvanize people in businesses right across the country,” she said. “By focusing on skills, data ethics and diversity, we can boost innovation and funding and put the UK at the forefront of this exciting and revolutionary sector.”

The new Office for AI was announced earlier this year and sits between the Department for Digital, Culture, Media and Sport and the Department for Business, Energy & Industrial Strategy. It has been tasked with overseeing the UK’s AI strategy and managing the industrial “grand challenge” on AI, which focuses on promoting data as a means to improving the detection of chronic diseases.

DeepMind’s health division is considered a pioneer in the field, and would be eligible for funding as part of the “grand challenge”. An Innovate UK spokesperson told NS Tech: “Each award Innovate UK makes is subject to strict eligibility requirements. We take due diligence very seriously and have in the past excluded applicants when conflicts of interest have arisen.”

DeepMind has won plaudits for its collaboration with Moorfields Eye Hospital on accelerating the detection of sight-loss conditions. But it has also attracted criticism for one aspect of its work in the health sector. A partnership with the Royal Free Hospital Trust in London prompted an investigation by the Information Commissioner’s Office in 2016. The ICO ruled that the partnership had initially failed to comply with data protection rules, but that the Royal Free, not DeepMind, was to blame for the violation.