With the enactment of the Artificial Intelligence (AI) Act in August of 2024, the European Union (EU) looks to lead the way in regards to the safe and thoughtful development and implementation of AI.
In the first part of this two-part series, we looked at the United States’ approach to AI regulation and found that while there is some work being done, the US still has work to do towards providing meaningful and effective laws and regulations that would prevent the misuse and abuse of AI. The EU’s AI Act looks to address many of these shortcomings by providing a well developed, clear, and concise framework for regulations on AI, while still allowing for enough space and flexibility for innovation to flourish. Though there are many similarities in the aims of the US and the EU, their levels of implementation are more different than they are similar.
The US has chosen to approach AI regulation in a relatively hands off manner at the federal level. Instead of tailor-made legislation, it interprets existing regulations in broader terms that allow for some level of AI oversight. This vacuum has made it necessary for many states to enact their own laws to prevent misuse and abuse. However, this fractious implementation lacks any cohesive vision despite the direction laid out by President Biden in his Executive Order on the Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence. While this order and the EU’s AI Act both call for safe, transparent, accountable AI that respects privacy, and is free from bias, the US fails to provide any regulatory framework to achieve this goal. The AI Act, however, is accompanied by a robust system of laws that is broken down into four major groupings: deployers, providers, importers, and distributors. There are also considerations within these categories for the level of risk that is presented in each use case. The video game industry, for instance, sees little to no serious regulations whereas the financial industry faces the highest level of oversight.
As one would expect, the differing approaches to AI regulation in the US and the EU have had, or will have, significantly different outcomes. The US’s laissez faire approach heavily favors the rapid development and implementation of AI technology. Developers in the US can move with relative freedom from the hindrance of regulations and other requirements. The obvious downside to this is that the regulatory framework that slows development in the EU also provides protections from misuses such as the improper handling of private data or potential abuses such as introducing bias into a selection system.
There is a reasonable fear that the US’s lack of regulations may result in major technology companies like Google, Microsoft, Amazon, and Apple shaping the future of the country’s AI industry to their benefit. It could also be argued, however, that the absence of regulation will encourage smaller companies to enter into the space. The costs of regulatory compliance could be seen as prohibitive to smaller companies looking to operate in the EU.
The difference in regulations could also have a significant impact on trade between the US and EU. While companies in the US may develop AI technology at a faster pace, they may struggle to adapt them into the stricter framework created by the EU. This should give organizations that develop their technology within the EU a significant advantage over foreign entities in regards to business within the EU itself.
Although the EU has a more fully developed system of AI regulations, its framework is still in its infancy. The AI Act has taken steps to address many of the issues outlined in President Biden’s executive order regarding bias and fairness, transparency and explainability, and balancing innovation with oversight. However, these emerging issues will need to be monitored and regulations will need to be developed or adjusted in order to adequately compensate. Adhering to these fundamental principles will ensure that AI technology is developed in a way that respects the rights of individuals as well as society as a whole.
The development and implementation of AI technology is moving at a furious pace. While the general aim of these advancements is to improve speed and efficiency, it is necessary to ensure that these benefits don’t come at the cost of individuals’ rights. In the near future, it is likely that individual states and regulatory agencies in the US will need to interpret old laws and develop new regulations to cobble together policies until such time as the federal government introduces its own comprehensive framework.
In the meantime, the EU looks to become a global leader in the field of AI regulation as it fine tunes the delicate balance between innovation and governance. By looking at google search trends in wake of the enactment of the AI Act, it is clear that the US will be watching very closely. However, for AI technology to truly flourish, countries need to develop cooperative agreements such as the work being done at the Council of Europe Framework Convention on Artificial Intelligence and Human Rights, Democracy, and the Rule of Law. Global cooperation will create an atmosphere where safe AI technology can be rapidly adopted from country to country, benefiting the world as a whole.