I started a new course last week to brush up on my AI ethics – the ‘Oxford AI Ethics, Regulation and Compliance Programme’ at Saïd Business School, University of Oxford.
It’s interesting to note which industries are represented in the 250+ people joining the course online, including staff from some of the big players in tech and social media, as well as legal, cyber security, healthcare (including the NHS), pharma, automotive, financial services, government, insurance and gambling industries.
We are currently on week 1, covering the most notable AI failures such as bias and discrimination, rights violations, misinformation, explainability/replication (or lack of), overreliance, and job displacement.
The course uses the EU Charter on Fundamental Rights as a key reference point for understanding AI ethics, and notes how we need to be wary of AI infringing individual rights, such as personal autonomy and mental health (something to note for those of us developing coaching chatbots). But it also suggests we reflect on the environmental footprint of AI through the same ethical lens, given how resource-intensive LLM training and deployment have become on a global level. Perhaps we need to think of this beyond values-signalling or organisational optics, and consider whether AI is also a threat to fundamental human rights at the environmental level?
I hope to share more discussion points from the course as I go through it!