EU regulators will introduce crash test systems designed exclusively for artificial intelligence

The European Union will introduce crash test systems designed exclusively for artificial intelligence. Before new inventions and technologies are introduced to the market, EU regulators will ensure they are safe to use through this process.

To ensure that innovations powered by artificial intelligence are safe before going on sale, the European Union established four permanent testing and experimental facilities across the continent. According to Bloomberg, the trade bloc invested about US$240 million (€220 million) in this project.

It will allow technology companies a space to test AI and robotics in various industries, including manufacturing, health care, agriculture and food, and cities, through the crash test systems and facilities that will be launched starting next year. Since technology is developing quickly daily, testing it is the only sensible course of action for the trading block.

Innovators are anticipated to provide new AI-powered tools to the market as "trustworthy" goods, according to EU Director for Artificial Intelligence and Digital Industry Lucilla Sioli, who announced at a launch event in Copenhagen. Sioli also called out misinformation as one of the dangers that AI poses to people.

European consumer organizations encouraged regulators to start looking at the possible dangers of generative AI, including ChatGPT. Consumer advocacy groups hope that by making this effort, existing laws will be enforced to protect consumers.

Ursula Pachl, deputy director general of the BEUC, voiced her worries about the technology, pointing out the possibility of deceit, manipulation, and harm to people. The statement made public claims that this will also aid in the systems' facilitation of false information, bias amplification, and fraudulent actions.

"We call on safety, data, and consumer protection authorities to start investigations now and not wait idly for all manner of consumer harm to have occurred before they take action," Pachl said. Authorities must enforce these regulations, which apply to all goods and services, AI-powered or not.

The head of the European Union initiated the AI Act before this endeavor. According to a story from The Guardian, this is the first significant attempt to govern the technology, which has already been in the works for two years. According to the risk they represent to consumers, AI systems are divided into four categories by the act: unacceptable risk, high risk, restricted risk, and low or no risk.

Technology that encourages manipulating people, encouraging risky behavior in children, social scoring, policing systems based on profiling, location, or past criminal behavior, and identification via biometric systems will eventually be outlawed once the trade block deems it to be an unacceptable risk.

After MEPs cast their votes in mid-June to advance this bill to pass an adjusted version of the draught put forth by the commission, the EU hopes to reach an agreement on the final draught by the end of this year. There currently needs to be trilateral discussions between the EU, the chairs of the AI committees in the EU Parliament, and the EU Council to advance the legislation.