Maria Maalouf
The European Parliament last week passed the Artificial Intelligence Act, which is viewed as the world’s first comprehensive law to define the legal aspects of AI. It provides for EU-wide regulations in reference to quality of data, transparency, human oversight and accountability. The law will apply to the providers and developers of all AI systems that are marketed or used within the EU, regardless of whether they are based inside or outside the bloc. Therefore, it will have implications for the use of AI in the US.
The act will impose fines on any company or group violating its rules. The exact amount will be a percentage of the global annual revenue that AI generates. The European AI law is being defended as not stifling innovation. Quite the opposite, in fact, as small and medium-sized enterprises will be adopting AI techniques. This will enhance innovation.
The European law on AI squares well with the main scope of AI, which is to achieve a competitive advantage in e-commerce. This will be gained in the processes of manufacturing, managing human resources, fostering customer relations and promoting advertisements and marketing. In other words, enterprises’ performance will be raised and this can increase profits, lower costs, save time, better collect data and reduce production and distribution errors, among other advantages.
The political history of the new act began in April 2021. It was the culmination of negotiations between the Council of the European Union and the European Parliament. Agreement on the text was reached in December 2023. Last week’s vote in the European Parliament turned that political consensus into law.
Most of its bylaws will enter into force two years after its introduction. However, any provision that has to do with the application of prohibited AI systems will be enforced within six months of its signing into law. In addition, any of its stipulations in reference to generative AI will be enforced after one year.
The act will abolish several AI applications that are considered to be a threat to the rights of citizens. These are related to biometric categorization systems that originate from sensitive characteristics. This will also deal with the practice of untargeted scraping of facial images by the internet or in CCTV. Such a practice aims to create a facial recognition database. These AI applications will be forbidden, in addition to the techniques that AI employs to control emotion recognition. These methods are considered manipulative of human behavior and exploit humans’ vulnerabilities.
The law is a direct response to the proposals submitted by average European citizens to the Conference on the Future of Europe in 2021-22. Many of their concerns centered on how to boost the European economic advantage in a few strategic economic sectors. They wanted to improve their ability to access information and data regardless of any particular trait or ethnic, political or religious affiliation.
The European AI Office will now be founded. It will function according to the European Commission’s rules. It will not conflict with the AI authorities and government bodies controlling the AI operations of individual member states. It will rather offer guidance on how European countries can best utilize AI and its capabilities. It promises not to discriminate against any European government or organization using this technology. However, political observers will not rule out a potential conflict involving this office and European entities since the implementation of AI involves complex processes and procedures and they can lead to many conflicts of interest among the people employing them.
Two different reactions have emerged regarding the new European AI Act. Human rights groups have condemned it.
The Access Now organization issued a statement saying: “The new AI Act is littered with concessions to industry lobbying, exemptions for the most dangerous uses of AI by law enforcement and migration authorities, and prohibitions so full of loopholes that they don’t actually ban some of the most dangerous uses of AI.”
On the other hand, the UN General Assembly is studying many draft texts as it seeks to announce a statement that can be a general declaration of humankind’s use of AI.
Finally, the European AI Act will have a great impact on any company doing business in the EU. Every American company must review its AI systems before they are used in Europe. This reviewing process could be costly and time-consuming. And American businesses cannot rely on US law because there is no federal legislation that allows or bans certain AI policies. The US government and private sector companies have no other option but to abide by the European law.
The only American alternative to this law would be to rely on state laws and regulations governing privacy in the US. But this would be a very short-sighted approach as far as AI laws in America are concerned. Therefore, the Europeans are far ahead of any other nation or group of nations in providing guarantees on how to use AI the right way.
