Mismanagement Risks Should Inform Climate Change AI Regulation

We cannot afford to allow non compliant actors in the AI sphere (that is, those which do not adhere to global citizenship principles) to frame our thinking about climate change artificial intelligence. It is difficult to overstate how damaging scandals relating to allegations of misused operation of AI are, at a time when global citizens need to assess the real potential of AI in dealing with the climate crisis. With COP26 in Glasgow close, we need to insist that regulation of climate change AI be initiated by a panel similar to the IPCC, the United Nations Intergovernmental Panel on Climate Change.

In general terms, revelations about organisations with global reach and power and how they use AI to adversely influence millions of people, some of whom are vulnerable, serve to warn us of risk. As case histories, they inform us about the rise of unregulated, unaccountable, and secretive use of AI technology.

The issue of global AI ethics effectively divides AI developers into “good” and “bad” global citizens and actors. The hacking community deals harshly with those who deviate from perceived internet derived ethical norms. However, if corporations ignore their responsibilities as global citizens by putting profit and growth ahead of accountability, transparency and privacy, we will continue to see revelations of apparent misuse of AI technology. These instances can inform the management and auditing of climate change AI operations in future.

​Climate change AI needs to be regulated so that it behaves in accordance with global citizenship principles. It needs to have a transparent structure, and it needs to answer to an independent governance body. Ideally, as a public/private hybrid, it will be non-profit.

​Big Tech is not the devil here, and neither is AI. At the current stage of its development, AI lacks the ability to mimic human intelligence to the point where it can begin to make political decisions or manipulate data to do destructive things. It is at times being mismanaged.

​AWS and Open AI are paving the way for ethical AI users by creating a model of open, democratised and thus safer, AI. Climate change AI is close, but it requires support from all of us, particularly, but not necessarily, the UN.

​A former Facebook employee, Frances Haugen, has emphasised at a United States Senate hearing, the need for algorithmic regulations, research transparency, and independent oversight in the AI driven social media space.

​The UN has embraced the concept of global citizenship, and needs to employ the global citizenship principles embodied in internet operation to enable ethical development of climate change AI in particular, but also AI use in general. These principles include universality and access, sustainability, equality, transparency of operation, independence from political interference, privacy of electronic communications, data protection, collaboration to guarantee infrastructure, and adherence to UN human rights agreements.

​Allegations of mismanaged AI should serve to encourage us to look towards regulation of AI, rather than putting off the problem for another day and another meeting. The climate change emergency means we need to move on the issue now, and climate change AI regulation may prove to be the catalyst for better AI use in every field.

​In balancing risk, the global risks of rapid climate change are far more significant than the risks of well-regulated climate change AI. COP26 needs to act.