Editorial: Who’s in control when AI decides?

The Ethical Adoption of AI, released recently by the Archdiocese of Malta, warns that “true progress is only possible when innovation is guided by trust and transparency”. It asked stakeholders to ensure that artificial intelligence “serves the human person and safeguards our common home”.

The position paper comes as we stand at a major shift in digitisation, perhaps reassured after the first few years that it will deliver the promised efficiencies and effectiveness – but perhaps not so sure about the cost in human terms.

The Church paper does not shy away from acknowledging the positive changes AI will bring but it does want more awareness on the fallout of the rapid changes we are facing, and it is calling on the government to introduce incentives for businesses to retrain and redeploy employees affected by automation. 

The paper’s emphasis on training is crucial. The archdiocese has provided AI literacy and critical skills training to more than 1,500 educators across its schools and is collaborating with the Malta Digital Innovation Authority to reach out to parents and vulnerable groups. 

It is hard to believe that, only a couple of years ago, teachers were concerned about children using ChatGPT to spare themselves the effort of homework. Schools quickly realised it was better to embrace the new technology than trying to contain it.

Teachers have embraced the inevitability of AI and, instead, are explaining the importance of using it wisely, cross-checking its references and using it as a background against which to grow their creativity, knowledge and curiosity.

AI has rapidly become indispensable and inevitable. But is it possible to control, to make sure – as the archdiocese says – that decisions are still taken by humans? Even if we do use it as a tool, surely our decisions are based on information it has processed, filtered and presented to us, which means it has already taken some of its own decisions!

Compare AI to a car, which was a tool for getting from Point A to Point B but which also evolved to become a weapon of death and destruction in recent years. Imagine how much more powerful AI is than a mere hunk of metal that stalled when you took your foot off the pedal. Imagine how much harder it will be to harness its power, especially when that power is not only available to those with good intentions.

The position of the archdiocese is based on stronger ethical standards and literacy efforts, using the term

“algorethics” to explain its vision. This is a “framework of transparency, accountability and inclusion where progress is measured by human wellbeing and social trust, not just GDP”.

The concept of well-being as opposed to economic growth is by no means new but it is important to understand its use in the context of AI.

The archdiocese is right to be concerned about AI’s impact on workers and social inequalities – but the ethical standards it is calling for are far from universal, in spite of efforts to legislate and control its abuse and exploitation, protecting users against manipulation in this era of fake news, propaganda and misinformation.

Given the new world order, global trade developments and shareholder pressure for profits, who is going to count the cost of what the Church warns against: AI-driven change that might “replace people or leave vulnerable communities behind”?

For decades we talked about a “brave new world”. We cannot say that we were not warned but the pace of change has left us unprepared and bewildered.

Read this position paper if you want to understand the “unintended consequences”.