AI: What, When, and How to Regulate? With Lessons from Estonia
The efficiency of modern data processing, backed by computerisation, poses many practical challenges to existing regulatory models in many sectors. Attempts to slow the speed of data exchanges potentially imperils the freedom of information upon which the modern digital economy and AI systems depend. Conversely, permitting them to remain unchecked ignores the risk of computerised systems creating runaway responses to each other with little opportunity for human intervention. It is clear that new rules and institutions will be required to govern processes which show little prospect of slowing down, much less stopping.
In considering whether and how specific sectoral guidelines should be introduced, some general lessons can be drawn from Estonia, a world leader in e-governance, which has had to grapple with such practical questions in relation to laws and regulations. While specialised legislation may appear attractive at first glance, the risk of obsolescence and lack of ongoing oversight suggests that such an approach should be treated with caution. Consideration must be given to the context within which AI systems are used; some legal frameworks can remain largely unchanged, whereas others may require amendments to address novel issues, including meta- legal implications such as the future of society and the role of humans.
This seminar is based on two articles published in the 2021 Singapore Academy of Law Journal Special Issue on Law and Technology.