The Future. The European Union has put forth the first draft law to regulate the development of AI in the West, taking a very comprehensive approach to mitigating the tech’s potential harm. The voting body hopes to pass the law’s final version by the end of this year, which could become the blueprint for governments worldwide.
The EU’s AI Act is an attempt to take control of AI… before it can take over us.
- It regulates how companies train their AI models, including publishing the copyrighted data used in the dataset and ensuring that the models can’t create illegal content.
- WSJ says the copyright provision could “give publishers and content creators a potential means to seek a share of profits when their works are used as source material for AI-generated content.”
- It also bars firms from collecting real-time biometric data in public spaces, aggregating data for facial recognition tools, or training “predictive” policing systems.
- Under the current draft, companies could be fined up to 7% of their global revenue for non-compliance.
While some in the tech sphere believe that the law — or any regulations, really — stifles innovation (Marc Andreessen of a16z), many others in the industry (everyone on this list and this list) believe regulations are quickly needed to keep AI from becoming dangerous.