Will we have a say in Ireland's artificial intelligence strategy?

Will we have a say in Ireland's artificial intelligence strategy?

Public accessibility and transparency of the process behind Ireland's AI policies and practices has been very poor to date, writes Dr Maria Murphy, Department of Law

Artificial intelligence (AI) is recognised as a crucial driver of technological innovation and economic growth. Many conveniences with which we are intimately familiar – from voice recognition on our smart devices to Netflix’s suggested shows – rely on AI processes.

AI is playing an increasingly pervasive role in managing our relationships with the external world. AI processes can assist in the shortlisting of interview candidates, the granting of credit, and the monitoring of suspicious activity on bank accounts. Public authorities are also increasingly using AI solutions in their interactions with ordinary individuals. Some jurisdictions have used AI processes in the criminal justice and judicial system.

While the use of AI creates great opportunities, important legal and ethical considerations from privacy to discrimination arise with its use. In acknowledgement of the benefits and potential costs of AI, the EU encouraged all Member States to adopt National AI Strategies by mid-2019. The European Commission’s Coordinated Plan on Artificial Intelligence supports ‘ethical, secure and cutting-edge AI’ and provides a strategic framework for the National Strategies.

The fact that Ireland has yet to publish an AI strategy seems surprising in light of Ireland’s positioning as a global technology hub and efforts to be branded as the ‘AI Island’. Indeed, some have commented that the ambition to become an island of AI is closer ‘to a dream than reality’. Notwithstanding delays, the Irish National AI Strategy is expected to be published in the first quarter of this year.

While the strategy cannot be judged in advance of its publication, there are issues with the process that should be of interest to everyone who lives in Ireland. A recent report by the Michael Dukakis Center for AI and Digital Policy considered the AI strategies of 30 countries and evaluated them on a number of key metrics.

Ireland would likely score quite well on some of the metrics due to its commitment to the OECD AI Principles, the Universal Declaration of Human Rights and EU data protection law. Unfortunately, even without seeing Ireland’s forthcoming strategy, one can predict that Ireland’s ranking would likely fall down on some very important points. In particular, it would be difficult to positively rate the public accessibility of Ireland’s AI policies and practices.

While other countries – such as Poland – have also yet to publish their AI strategies, it is common to publish draft policies and other preceding documents in order to keep the public up to date on the development and direction of government policy. Indeed, Poland provides an AI roadmap where the public can readily access information regarding the strategy’s progress and milestones. Poland has also published a draft version of its strategy online. 

In spite of the ongoing work of the Department of Enterprise, Trade and Employment in the development of an Irish strategy, documents concerning the process have not been published online so far. In order to encourage engagement and transparency, the publishing of material online is clearly best practice. An Irish resident should not need to resort to a Freedom of Information request to obtain the type of information that is widely available to the public in other countries.
It is positive that a public consultation process – "designed to better understand the views of the public on the opportunities, enablers and challenges for AI in Ireland" – was launched. Yet the consultation process closed in November 2019 and the public submissions have not been shared online. A central repository of submissions would be a valuable resource for the public discourse on AI, particularly in light of the ongoing delay in the publication of the report itself.

Transparency is a vital democratic value and it is particularly important in the context of AI. While we may all be on familiar terms with Alexa and Siri, we may not realise the processes that underlie their functioning. Surveys show that people are often unaware of how and when AI-powered automated decisions are being made that affect their lives in very real ways. Considering the transformative potential of AI in our society, it is imperative that the public are a part of the conversation and not simply the audience concerning our AI strategy.

There is clear scope for a broad public conversation and greater inclusion of all stakeholders in the discussion. Indeed, the need for transparency in government policy concerning AI will not cease following the publication of the strategy. As the technology and its uses develop and evolve, it is imperative to keep the public informed and involved as active participants in the collective choosing of a direction for the future of AI in Ireland.