Hi friends! In today’s article we will discuss What is the EU’s New AI Act and other related issues. Today Artificial Intelligence (AI) is all the rage in the whole world.
AI is being used in almost every field and efforts are being made to enhance its work quality. But, we must not forget that AI is a tool and like any other tool, it can and is used for both good and bad reasons.
Artificial Intelligence (AI), as amazing as it was in the beginning, has now become very dangerous. Many countries of the world including America and India are troubled by AI.
AI is beneficial but it is mostly being used for wrong things. How the power of the AI can be used for evil purposes can be seen in the Deepfake and Clearfake incidents that have taken place across the world.
All the world has been talking about regulating it for a long time but till now no law has been made in this regard. Therefore, efforts are being made to bring AI within the ambit of the law to control it.
The first step in this has been taken by the European Union (EU) which has agreed to create an Act against AI.
If the EU makes the AI Act, it will be the first time in the world that efforts will be made to bring AI within the ambit of law.
What is the EU’s New AI Act?
The EU gave its consent to make a law related to this last Friday. This act will be called the AI Act which will control the rapid development and risks in the field of AI.
This law bans harmful AI practices that involve people’s safety, livelihood and rights.
The proposal regarding the law against AI was given in the year 2021 itself.
European Parliament President Roberta Metsola said in a press conference that the law was necessary to prevent the misuse of AI which is rapidly increasing.
The AI Act is a legal framework governing the sale and use of artificial intelligence in the European Union.
Its official goal is to establish uniform standards for AI systems among EU member states, thereby guaranteeing the smooth operation of the EU single market.
You may also like:
What Does the AI Act Cover?
The EU states that the purpose of the Act is to safeguard against high-risk AI the principles of democracy, environmental sustainability, rule of law, and fundamental rights.
Simultaneously, it will endeavor to guarantee that it encourages innovation in Europe and assists the continent in emerging as a the leader in this domain.
AI systems that are used, deployed, or put on the market in the EU are covered by the AI Act. It follows that this applies not only to EU developers and deployers, but also to international vendors who sell to or otherwise make their system or its output available to EU users.
There are three exceptions:
- Artificial intelligence (AI) systems created specifically for military use, with the potential to be expanded for wider defense and national security applications, subject to negotiations;
- Artificial intelligence created and applied to scientific research; and,
- AI systems and components that are free and open source (the term is not yet well defined), with the exception of foundation models, which are covered below.
What are the Unacceptable and High Risk Systems for EU AI Act?
Risk can be categorized into four categories: minimal/none, high, limited, and unacceptable. The main focus of the AI act is unacceptable and high risk categories.
Unacceptable Risk:
- AI systems that fall under the unacceptable risk umbrella are strictly prohibited.
- The three proposals reached a consensus that risk systems that are deemed unacceptable are those that possess a high potential for manipulation, either by means of subliminal messages and stimuli or by taking advantage of vulnerabilities such as age, disability, or socioeconomic status.
- Additionally prohibited are AI systems used for social scoring—a term used to characterize the assessment and handling of individuals based on their social behavior.
- Along with other biometrics and law enforcement use cases, the European Parliament also plans to outlaw real-time remote biometric identification in public areas, such as live facial recognition systems.
Applications that are Banned:
The co-legislators decided to prohibit the following because they understood how some applications of AI could endanger citizens’ rights and democracy:
- Systems for biometric classification based on sensitive attributes (such as race, sexual orientation, religion, or political views);
- Untargeted internet or CCTV video scraping of faces in order to build databases for facial recognition;
- Acknowledgment of emotions in the workplace and in educational settings;
- Social scoring determined by an individual’s social behavior or personal traits;
- Artificial intelligence systems that control human behavior to evade free will;
- AI is used to take advantage of people’s weaknesses (caused by their age, disability, social or economic situation).
High Risk Systems:
Systems that pose a high risk will be carefully regulated. Followings are believed to fall into the following eight broad categories:
- Utilizing biometrics
- Essential infrastructure
- Instruction and career development
- Workplace employment, worker management, and self-employment accessibility
- Obtaining necessary services
- Law enforcement
- Management of migration, asylum, and border controls
- judicial administration and democratic procedures
Provision for Penalties:
Penalties for noncompliance can range from €7.5 million ($8 million), or 1.5% of turnover, to €35 million, or 7% of a company’s worldwide turnover.
As per EU, the extent of the violation and the size of the company will determine all penalties.
Efforts Being Made by Other Countries:
China has imposed limitations on specific AI technologies, while Japan is developing nonbinding guidelines for the technology.
According to Britain, the laws in place are sufficient to regulate technology.
Saudi Arabia and the United Arab Emirates are spending funds on AI research.
Other countries including India are also working on it to control misusage of AI.
Conclusion:
The efforts being made by the EU to bring the AI within the ambit of law are excellent. This is a source of inspiration for the whole world.
That’s all my friend for now! I hope this article will be valuable for you. I will love your comments and feedback. You can reach me through the emails given below.
Yours
Abhijit Ranjan
1 thought on “What is the EU’s New AI Act? Big Update of 2023”