AI Regulation in the European Union

Gave a webinar today on AI Regulation in the European Union. The EU has some onerous draft regulations which look likely to become law soon

Fines for non-compliance are 30M euro or 6% of annual revenue – that should get the attention of boards around the world (the Europeans don’t muck around when setting penalties)

Here is what I took the first 10 minutes to say:

– Want to sell a tech business, start an AI business. Since 2010 there have been over 1,000 AI companies acquired by 800+ acquirers. Mostly corporates looking to leverage AI without building it from scratch. All sectors are buying inc Energy, Health & Retail

– AI’s most popular use cases are:
– Recruitment
– Customer Service (no one can stop the rise of the Chatbots)
– Lending & Credit Scoring (computer say no)
– Portfolio Management & Trading
– Systems & Optimisation
– Fraud Detection & IT Security

There is not much happening around the world in AI regulation (see graphic) except the EU where they have really given it a good go. There are some global regulatory themes emerging:
– Accountability (human oversight)
– Treating customers fairly
– Transparency (letting people know they are dealing with a Bot)
– Operational resilience of the AI system
– Obligations when outsourcing AI

The EU AI Regs will, like GDPR, impact any organisations with any link to the EU and the EU AI Regs will be at least as onerous to comply with as GDPR

Here is what I took the remaining 30 minutes to say:

THE EU AI REGULATIONS ARE REALLY COMPLICATED & COMPLIANCE IS GOING TO BE REALLY HARD