The Artificial Intelligence Act

Introduction

The European Parliament adopted the regulation (‘’AI Act’’) on Artificial Intelligence (AI), bringing it one step closer to its potential transformation into law. Upon approval by the Council, this Act will be the first-ever regulation dedicated to AI.

What is the AI Act?

The draft legislation includes comprehensive rules and definitions about AI. It covers critical aspects such as privacy, data security, and ethical usage while ensuring responsible and accountable practices in AI deployment.

What is the background of the AI Act?

Artificial intelligence (AI) and its wide-ranging applications have revolutionized various sectors, from creative activities like writing poetry to practical tasks such as contract drafting. As AI continues to advance, its centrality in our lives grows, significantly impacting our daily routines and interactions. However, it is critical to beware that AI's impact is not without any risk.

The increasing popularity of AI has led to global concerns regarding privacy, intellectual property rights, and plagiarism. These concerns have emerged as AI technologies become more prevalent in society. In this context, lots of experts along with the CEO of the OpenAI and Bill Gates, have declared an alarming statement about the potential dangers associated with AI:

‘’Mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war.’’

In response to the concerns regarding privacy, intellectual property, and plagiarism, the European Parliament has taken proactive measures to regulate these issues and adopted the AI Act.

By adopting this regulation, European Parliament aims to establish clear rules that ensure transparent, traceable, non-discriminatory, and environmentally friendly use of AI, safeguarding privacy, protecting intellectual property rights, and mitigating the risks associated with plagiarism.

What are the provisions of the AI Act?

The AI Act uses a risk-based approach to regulate AI, categorizing into four risk levels and each category comes with specific requirements to be followed.

The AI Act contains a comprehensive set of regulations to prohibit certain artificial intelligence practices. The AI Act prohibits the use of AI in biometric identification and categorization systems. It also forbids the deployment of AI systems that employ manipulative techniques. Additionally, the AI Act prohibits AI systems that exploit people's vulnerabilities and those that scores individuals based on their behaviors, economic or social class. Predictive policing systems based on profiling, location, or past criminal behavior are also prohibited under the Act. Furthermore, the use of emotion recognition systems in law enforcement, border management, workplace, and educational institutions are not allowed.

Under the Act, AI systems are required to comply with transparency obligations regarding the data used during their training. In addition, generative AI systems must indicate that they are the creator. Also, the data they trained on must not violate the copyright law.

What is the scope of the AI Act?

The provisions of the AI Act apply to users of AI systems that are located in the EU, operators of AI systems that are established or located in the EU, and providers, users, and operators that are established or located in a third country if the result produced by the system is to be used in the EU.

What are the penalties for non-compliance with the AI Act?

Member States are required to establish penalty rules to prevent any violations regarding the AI Act. According to the AI Act, these penalty rules must be effective and proportionate, taking into consideration the interests of small-scale providers, start-ups, and their economic viability.

Infringements related to the above-mentioned prohibited artificial practices (article 5) and non-compliance with data governance provisions (article 10) may result in administrative fines up to €30,000,000 or up to 6% of the offender's worldwide annual turnover, whichever is higher. Infringements related other than articles 5 and 10 may lead to administrative fines up to €20,000,000 or up to 4% of the offender's worldwide annual turnover, whichever is higher.

Providing false or misleading information in response to official requests from relevant authorities may result in administrative fines up to €10,000,000 or up to 2% of the offender's worldwide annual turnover, whichever is higher.

It is important to note that the exact amount of the administrative fine depends on the specific circumstances of each case, including the severity and consequences of the infringement. The size and market share of the offender should also be taken into account. Furthermore, if the offender has previously been subject to an administrative fine, this will also be considered.

What are the critics of the AI Act?

Critics claim that the AI Act would potentially hinder the development of AI by slowing down the fast advancement and application of AI technology. The Act's broad definition of AI is criticized by experts for lacking clarification and specificity. Due to this ambiguity, experts say, it may be difficult to consistently enforce the rules and they may be interpreted differently.

Conclusion

The AI Act represents a significant step in regulating AI and aims to ensure transparency, accountability, and ethical use. It should also be noted that, The Act is currently not enacted into law yet. Stay updated on the status of the legislation for the most accurate and current information.

Autor: Gürkan Erdebil
Autor: Müge Şengönül