By Christine Choong
Introduction
Many of you who are reading this might be thinking about ChatGPT. No, this does not only include artificial intelligence (“AI”) tools like ChatGPT. Artificial intelligence already permeates our lives in some way or another, even in other tools like Grammarly or even how TikTok recommends videos on your ‘For You Page’. So how do governments take steps to regulate this new technology and who has done it already?
What is the EU AI Act?
The EU AI Act, introduced by the European Commission in April 2021, is the first big attempt to regulate AI within the European Union. The main goal is to set clear rules for how AI should be developed and maintained while encouraging simplicity for small and medium-sized enterprises (“SMEs”) businesses. The Act outlines what AI developers and users need to do to comply with these rules, focusing on different levels of risk based on how the AI is being used. The act aims to categorise AI applications according to their risk levels, to ensure that AI is applied harmoniously and ethically in the marketplace.
The AI Act divides AI systems into four categories of risk: ‘Unacceptable risk’, ‘Limited risk’, ‘High risk’, and ‘Minimal risk’.
The ‘Unacceptable risk’ will be banned entirely from the EU market. This includes AI programs which manipulate human behaviour or adhere to vulnerable groups; with examples being, voice-activated toys that encourage kids to engage in dangerous activities, and AI mechanisms used for social scoring or biometric identification, such as facial recognition technology.
For AI systems considered to have limited risk, the Act asks for more transparency. This applies to AI used in products like toys, medical devices, or cars. AI used in education, worker management, law enforcement, and migration control must also be registered in an EU database. Generative AI, like ChatGPT, will have specific rules, such as making it clear when AI-generated content, ensuring that illegal content isn’t created, and showing what copyrighted data was used for training the AI.
Most AI systems fall under the minimal risk category, which means there aren’t many restrictions. This includes everyday applications like AI-powered video games or spam filters. Since the majority of AI currently used in the EU is low-risk, these systems can be used pretty freely without worrying too much about regulation.
How Does the AI Act Affect British Businesses?
For British businesses, the EU AI Act has a substantial impact, especially if they want to continue doing business in the EU. Even though the UK is no longer part of the EU after Brexit, British companies offering AI solutions will still need to follow the EU’s rules if they want access to the European market. This is true for industries that are involved in healthcare and finance, where AI systems are considered high-risk. If British companies don’t meet the EU’s standards, they could end up being unable to operate in the EU, which would likely lead to financial losses. Despite Brexit, the EU remains an important trading partner for the UK — about 42% of UK exports went to the EU in 2023 — so it’s still crucial for British businesses to maintain good trade relations.
Additionally, British businesses will also need to pay attention to EU data and privacy laws when using AI. The EU AI Act will work alongside the General Data Protection Regulation (“GDPR”), which means companies need to make sure that AI systems handle personal data properly. If they don’t, they could face fines or risk damaging their reputation.
Complying with the EU AI Act might also bring some challenges for British businesses, especially when it comes to innovation. Adapting to these new rules could mean spending more money on things like transparency, documentation, and testing to make sure their AI systems follow the regulations. While this might increase costs, it could also give British businesses a chance to show their commitment to responsible AI use, which could help them stand out in a competitive market.
Although the UK no longer has any obligations to comply with EU laws, British businesses that want to stay active in the EU will need to keep up with these new AI regulations. If they don’t, they risk being excluded from the EU market, which could have a big financial impact.
Conclusion
In summary, the EU AI Act introduces a risk-based framework for AI, with different requirements based on how risky the application is. British businesses will need to follow these rules if they want to stay in the EU market, which could mean adjusting their practices and investing in compliance. However, this also gives them a chance to position themselves as leaders in ethical AI, both within the EU and beyond.
留言