In careful stages, Europe has endorsed and built a vision for regulation around artificial intelligence. 2023 ended with a consensus reached by the EU Council states and the European Parliament during a three-day ‘marathon’ of talks to agree on the first rules of the AI Act, due to take effect in at least 2 years.
The European Parliament which is at the forefront of overseeing responsible AI innovation across Europe will vote on the Act’s principles in 2025.
Carme Artigas, Spanish secretary of state for digitalisation and artificial intelligence, said the agreement addressed “a global challenge in a fast-evolving technological environment” which would encourage investment and further expansion of AI in Europe.
AI Systems under the Act will be categorised in tiers in terms of the ‘risk-based approach’ which should be taken with higher risk determining more limitations. The aim of creating a ‘gold standard’ for AI regulation is still going be a heavy task with the emergence of newer technology problems such as deepfakes morphing into new challenges. Meanwhile, there is more known around the concerns of facial recognition and action being taken by solution providers to combat inherent threats to advance comprehensive laws for some aspects of AI.
Nonetheless, the deal is said to strike clarity on ‘historic’ rules over the use of high-impact general-purpose AI models and extend the prohibitions to keep up with the technology landscape impacting our societies and economies. The opportunities that AI provisions in policing will still be leveraged and balanced with stricter rules against ‘unacceptable’ risk.
The provisional agreement bans cognitive behavioural manipulation, the untargeted scraping of facial images from the internet or CCTV footage, emotion recognition in the workplace and educational institutions, social scoring, biometric categorisation to infer sensitive data, such as sexual orientation or religious beliefs, and some cases of predictive policing for individuals
