After COVID-19: what we’ve learned about the ethical application of AI
Richard Hacking, Head of UK&I AI Lab
Richard is the Head of Atos’s Automation & Intelligence Practice and a member of Atos’ Scientific and Expert Communities. Richard has over 20 years’ experience bridging the gap between business and IT. He has undertaken Programme & Account Management roles specialising in cloud and data intelligence domains which has included being the Programme Director for London 2012 Olympics. Richard is also Head of Atos’s UK&I AI Lab working with clients to discover how intelligent automation can benefit their organisation, design solutions and define their roadmaps to implement and scale solutions.
The response to the COVID-19 pandemic has often been characterised as a challenge of human ingenuity – from studies to identify how the virus spreads and replicates, to trials dedicated to developing new drugs and vaccines.
In time, however, this Herculean human effort may also be viewed as a turning point in the way we deploy artificial intelligence to tackle society’s biggest challenges and understand the ethical implications of that deployment.
Drawing the ethical line
By forcing the prioritisation of medical need above all else, the pandemic has inadvertently provided a test bed for AI and machine learning to be deployed in ways that may not otherwise have been possible.
It has brought to the fore privacy concerns over using AI to track patients’ conditions through apps and analysing vast sets of personal data, for example, and weighed those concerns against the urgent public health need. This has magnified the crucial question of where the ethical line should be drawn between the greater public good on the one hand and individuals’ right to privacy on the other.
Unleashing AI’s full power
Aside from analysis of patient data, AI is also being used to dramatically compress the timescales for research – from modelling viral transmission to screening candidate treatments. It is even helping researchers rapidly sift through the vast amount of scientific literature being produced on COVID-19 to find the most relevant new studies.
The case for unleashing the full power of AI to combat the public health threat is strong – but the path has not always been smooth because of the very valid ethical questions that this raises.
Bringing people onboard
For AI to be applied to its full potential, it first must be palatable. AI is some of the most powerful technology humanity will ever develop, and it’s beyond doubt that is has helped the global research community respond more quickly and intelligently to the threat of COVID-19.
But this will always be balanced against rightful enquiry into the potentially far-reaching consequences of its deployment – from privacy to potential bias in favour of certain groups or demographics; its impact on the human workforce; and the issue of who is accountable for decisions made or recommended by machines.
Using the experience of COVID-19 for the greater long-term good
We are at a critical time, where questions about the appropriate use and scope of AI are still to be fully considered – the experience of recent months has provided a test case into these questions offering penetrating insights.
Policymakers must absorb the lessons of this experience and apply them in the long-term, establishing the frameworks and approaches which will work in tandem with those who will design and deliver our future world.
As the technology sector, we too must strike the balance between advocating the best possible use of technology for the greater good, and providing confidence that when we do so, appropriate controls are being applied.
By working together, we can ensure the benefits of the technological revolution are shared by everyone in society.
Head of AI Lab, UK & Ireland