Trust, transparency must form pillars of Singapore's AI success

Artificial intelligence (AI) is unlocking transformative opportunity around the globe. From diagnosing disease to directing traffic on our streets, AI is changing the way we live and work. So how do we balance this emerging reality with the fundamental responsibilities of government?

It's clear that AI is an exciting, but also intimidating opportunity. Even aside from Hollywood stories of murderous robots ingrained in our collective subconscious, questions of appropriate data use and privacy are enough to raise warning flags for many.

In Boston Consulting Group's The Citizen's Perspective on the Use of AI in Government, we explore these issues on a global scale. With insight from over 14,000 respondents in 30 countries around the world, it's clear that international opinion remains cautious about widespread AI use in society. In addressing that challenge I would argue that two core principles of government remain central to an effective approach - that of trust and transparency.

What's the attitude to AI in Singapore?

There is a broad correlation between trust in government and the net positive perception of AI, with Singapore holding a relatively positive position alongside the likes of Hong Kong, France, and Japan.

An interesting element of global opinion is the differentiated views across generations. Internationally, 57 per cent of those aged 18 to 34 are in favor of AI use by government, and only 21 per cent against, while those over the age of 60 had a net positive support of just 19 per cent.

While this pattern is broadly mirrored in Singapore, we actually find that those aged 35-49 are the positive cheerleaders of AI in the country, with 36 per cent strongly agreeing with use of AI by government. Equally intriguing is the role of the 50-to-59 age group, the most hesitant group, with 13 per cent strongly disagreeing with its use.

On the ground, public perception of AI and data use in Singapore has suffered some troubling setbacks in recent years. Widely publicised data breaches have provided high-profile cases of the perils of our data-driven world, not least the emotional resonance of hackers gaining access to private health records at SingHealth. Recent revelations around the leak of blood donor records have triggered a serious review of data management throughout government.

The nature of this challenge is why deepening trust and enhancing transparency are so vital in supporting our emerging AI opportunity.

The question of trust in AI adoption

Trust is fundamental to a positive economy - trust in institutions, in banks, in each other. That same trust must form the foundation of positive AI adoption by government.

Privacy and data handling is clearly an important element of this debate. The nature of the SingHealth hack is a particularly acute invasion of personal information, and highlights a significant concern revealed in our research about the role of technology in such highly emotional and ethical realms. Put simply, there's a very clear dividing line between what people trust AI to do, and what they don't.

In operational environments perceived to be driven by objective statistics, citizens are broadly positive about the use of AI. Among Singaporean respondents, 44 per cent strongly agreed with the use of AI in traffic optimisation, 42 per cent in machinery maintenance, and 39 per cent in tax assessments. Now compare that to areas of implementation perceived to involve ethical or emotional decisions.

The survey showed 30 per cent of Singaporean respondents strongly disagreed with the use of AI to assess guilt or innocence in a trial, 17 per cent strongly disagreed with AI making parole board decisions, and 15 per cent were strongly against AI making recommendations on medical treatment. Overall, 28 per cent of Singaporeans were strongly concerned that ethical issues had not been resolved.

The question of jobs is another crucial area of concern. While some analysts champion AI as the next great revolution in job creation, others loudly proclaim that millions are about to suffer from AI-driven unemployment. The average citizen is stuck in the middle with only their concerns for company. BCG's report found that 44 per cent of respondents in Singapore had strong concerns about AI resulting in lack of work in the future, 43 per cent strongly agreed they were concerned about the broader impact on jobs, and 39 per cent strongly agreed with government regulating AI to protect jobs.

Building trust with transparency

In the face of these key challenges, government must work hard to demonstrate trust in both AI, and their management of its implementation. That means the benefits, challenges, and operational realities of AI must be showcased in a clear and transparent way; 27 per cent of those asked were strongly concerned about the potential lack of transparency in AI decision making.

Singapore has taken encouraging steps as a world leader of AI adoption in government. The Smart Nation and Digital Government Group has already generated benefits in using AI to tackle corruption in procurement, developing smart homes, and supporting assisted living amongst other applications.

The issues of ethics and transparency are also being tackled. The framework for ethical use of AI represents a clear attempt to openly address questions of trust and morality. The Public Sector (Governance) Bill of January 2018 takes important steps to formalize agency data-sharing, data-handling, and criminalizes data-related offenses. The recently announced review of data handling could also provide a much-needed boost to public trust.

Government must also address more nebulous fears such as potential job losses from AI. Transparency in addressing this will be vital for building trust, as will be providing frameworks to support talent to reskill for this emerging future. Plans for an inter-agency AI task force show positive steps in this regard, with AI talent development a key part of the proposal.

The only way is forward with AI

The AI genie is out of its bottle, and no amount of wishing it were otherwise will turn back the tide of AI innovation.

AI is undoubtedly going to open up transformative global opportunity. Only by building on these twin pillars of trust and transparency can governments steer this change in the best possible way. That means being transparent about the challenges, clear about the steps that are being taken, and committing to tackling the serious ethical and operational questions which are sure to arise.

Singapore should be proud of the steps it has taken to date. Promising plans for the future speak to further consideration in charting the course ahead. Yet with a technology such as AI, as the opportunity evolves, so too must our attitude towards it. By building trust in the public, and transparency in its actions, the government is best placed to ensure that change is for the better.