A Path from Ethical Principles to Law

The Issue

The purpose of ethics and the law are often distinct yet the EU is on a path to turn ethical principles into legal rules. Is this the right approach?
Recently the European Parliament’s Legal Affairs Committee released an AI report which outlines an approach for Europe to turn its AI ethical principles into a regulatory framework.

The key components of the proposed regulatory framework include:

A set of ethical principles, which are drawn from the work of the EU’s High Level Expert Group on AI;

Establishment of national supervisory authorities within Member states to assess the risk and monitor if any AI or associated technologies are at high risk of breaching the ethical principles;

Establishment of an EU Agency on AI to supervise the application of the proposed regulation, ensure harmonisation across member states as well as developing a European certificate of compliance with ethical principles.

Much work and discussion is still to occur to make this a reality but in the end it will be up to the national supervisory authorities to draw the line in the sand on ethical breaches. The words used in the proposed legislation are:

(9) Any artificial intelligence, robotics and related technologies, including the software,
algorithms and data used or produced by such technologies, which entails a high risk
of breaching the principles of safety, transparency, accountability, non-bias or nondiscrimination, social responsibility and gender balance, environmental friendliness and sustainability, privacy and governance, should be considered high-risk from a compliance with ethical principles perspective where that is the conclusion of an impartial, regulated and external risk assessment by the national supervisory authority [my highlights].

(10) Notwithstanding the risk assessment carried out in relation to compliance with ethical principles, artificial intelligence, robotics and related technologies, including the software, algorithms and data used or produced by such technologies, should always be assessed as to their risk on the basis of objective criteria and in line with relevant sector-specific legislation applicable in different fields [my highlights] such as those of health, transport, employment, justice and home affairs, media, education and culture.

Perhaps the element of this proposed legislation that stood out most was contained in the Explanatory Statement. It references the movie Blade Runner as well as Isaac Asimov’s Foundation series. Here is a verbatim extract:
Explanatory Statement In 1982 film ‘Blade Runner’, Rachael, a ‘replicant’ who works for a company that manufactures other ‘replicants’ – sentient humanoid robots – says to Deckard, a bounty hunter who makes his living eliminating rogue replicants: – ‘It seems you feel our work is not a benefit to the public.’ Deckard replies: – ‘Replicants are like any other machine – they’re either a benefit or a hazard. If they’re a benefit, it’s not my problem.’ …. The EU’s role in establishing a legal framework When public administrations address this phenomenon we cannot, however, adopt Deckard’s professional cynicism. For the European Parliament it is just as important to harness these technologies’ potential benefits for Europe’s well-being and competitiveness as it is to monitor their inherent risks, or to pre-empt the consequences of the any of those risks actually manifesting itself……

Our View

The European Union is on a clear path to AI regulation and this paper progresses that intent, but only marginally. The real work will still need to be done by national governments in setting up relevant AI supervisory agencies, developing processes to undertake risk assessments and monitor AI ethical compliance. Additionally, this framework requires an objective set of criteria be developed in making AI risk assessments while also taking into account industry-specific factors. There is still lots of work to do.

Open questions:

  • Are works of popular science fiction the appropriate inspiration for creating the right legal and ethical frameworks for AI? If so, perhaps we should systematically tap into our creative arts communities to start to imagine and explore plausible AI futures?
  • When creating legal rules, are ethical principles the best starting point when legal rules describe what we must do while ethical principles describe what we ought to do?