Nicosia, 03 December 2020
EY global study finds disconnect on ethical artificial intelligence (AI) priorities between public and private sector
- Discrepancies exist between companies and policymakers on future of AI governance
- Both groups show potential blind spots, posing risks
EY releases findings from a global survey finding significant differences in how the public and private sectors view the future of ethics, governance, privacy, policy and regulation of artificial intelligence (AI) technologies. According to the Bridging AI’s trust gaps report, developed in collaboration with The Future Society, AI discrepancies exist in four key areas: fairness and avoiding bias, innovation, data access and privacy and data rights.
In the survey of 71 policymakers and more than 280 global organizations, respondents ranked ethical principles by importance for 12 different AI use cases, and sentiment was measured around the risk and regulation of AI.
Policymakers align around specific priorities, while private sector lacks consensus
Policymakers’ responses show widespread agreement on the ethical principles most relevant for different applications of AI. For example, on the use of AI for facial recognition policymakers rated “fairness and avoiding bias” and “privacy and data rights” as the two top concerns by a wide margin. Yet private sector priorities on the same question were relatively undifferentiated. In fact, the private sector responses across use cases and principles were more evenly distributed, with narrow margins defining the top choices. And the private sector’s top choices were principles prioritized by existing regulations, such as GDPR, rather than emerging issues such as fairness and non-discrimination.
Disagreement about the future direction of governance poses risks
While both policymakers and companies agree a multi-stakeholder approach is needed to guide the direction of AI governance, results show disagreement on what form it will take: 38% of organizations surveyed expect the private sector to lead a multi-stakeholder framework, and only 6% of policymakers agree. This disconnect poses potential challenges for both groups in driving governance forward, and it also presents market and regulatory risks for companies developing AI products while governance approaches are still under discussion.
Overcoming differences throughcollaboration
The survey results found that each stakeholder group has blind spots when it comes to the implementation of ethical AI, with 69% of companies agreeing that regulators understand the complexities of AI technologies and business challenges, while 66% of policymakers disagreed.
These findings suggest greater collaboration between both groups will be critical to overcome knowledge gaps. Policymakers should take a consultative and deliberate approach with input from the private sector, particularly on technical and business complexities for which policymakers lack expertise. Similarly, the private sector should work to reach consensus around AI governance principles, so the regulation requirements of both parties are taken into account.
Commenting on the survey findings, Charalambos Constantinou, Partner and Head of Consulting of EY Cyprus, said: “AI is transforming businesses, the global economy and indeed our society. Yet, to establish public trust in this technology, a number of ethics, governance and privacy issues need to be urgently addressed. This will require a much closer alignment between decision makers and companies, so that the proper framework can be created that will allow the technology to develop and critical applications to be adopted.”
To access the Bridging AI’s trust gaps report, click here.