The European Union are a step closer to adopting the proposed Artificial Intelligence Act (“EU AI Act”), which is a legal framework by which AI protocols will be regulated across the 27 member states that make up the EU.
European Commission Votes on Landmark Legislation
Today, being February 13th, the European Parliament’s Committee is expected to vote on the proposal. This comes after the ambassadors of all the member states unanimously approved the AI Act, including France, which had opposed the rules for reportedly more than seven months.
After the voting today, the Act progresses to the European Parliament for voting and then formal adoption at the ministerial level. If this happens, the Act could indeed enter into force before the summer. Some parts of the Act may then apply as early as 2024, with full application likely in 2026.
EU’s AI Act is a key development that will shape the development and usage of AI systems in Europe going forward. It was introduced by the European Commission three years ago, proposing different parameters by which all AI systems shall be regulated across diverse sectors.
EU Balances AI Regulation
One noteworthy thing about the Act is the approach by which it seeks to regulate AI systems. Rather than a one-size-fits-all approach, the Act seeks to regulate AI systems based on their level of risk to fundamental rights.
The Australian government is also taking a similar approach to regulating AI in their nation. So, basically, Australia and the EU don’t blanket-regulate all AI applications but rather focus on those deemed high-risk, such as those used in healthcare, finance, or critical infrastructure.
Applications deemed to be high-risk will face stricter requirements for transparency, safety, and fairness. However, those lower-risk applications, like simple chatbots, will be subjected to less stringent regulations.
To many, the EU’s regulatory approach is a balance between innovation and safety. Overly restrictive regulations could stifle progress, while insufficient safeguards could harm individuals and society.
It’s also worth mentioning that the rules are not all figured out. For instance, there is no clarity on the risk level of AI systems will be assessed.
“Many of the new AI rules remain unclear and could slow down the development and roll-out of innovative AI applications in Europe,” CCIA Europe’s Senior Policy Manager Boniface de Champris said. “The Act’s proper implementation will therefore be crucial to ensuring that AI rules do not overburden companies in their quest to innovate and compete in a thriving, highly dynamic market.”
