European Union passes landmark AI bill as Congress debates first steps to set guardrails – The National Desk

4 minutes, 10 seconds Read

The European Union passed its landmark artificial intelligence regulation, setting the tone for creating guardrails for a rapidly advancing technology that has enveloped tech investors and governments around the world seeking to limit the potential downsides.

Under the Artificial Intelligence Act that was passed by overwhelming margins in the European Parliament, AI systems face more scrutiny based on how risky they are deemed to be. In high-risk uses like electrical networks or medical devices, companies would have to comply with stricter requirements and provide clear information to their users.

The law also bans the use of AI altogether in some applications, like social scoring systems governing how people behave and police scanning faces using AI-powered biometric identification systems with narrow exceptions.

AI-generated photos and videos, known as “deepfakes,” are also required to be clearly labeled under the bill, which addresses a widespread concern about the spread of misinformation powered by artificial intelligence. Dealing with deepfakes has been a priority in countries around the world as many of them are set to hold elections this year, the first after AI’s use capability and popularity have exploded since the introduction of OpenAI’s ChatGPT.

There are still several steps to go before the law is implemented in the 27-nation bloc of EU member countries. Exactly how the law will take effect and be implemented is also a bit of an unknown at this point due to AI’s advancing capabilities.

“We have to see how changes in the technology that are almost certainly going to happen in the next year, two years, five years, etc., will interact with this bill, and by extension, any potential laws that we’re considering here in the United States,” said Matt Mittelsteadt, a research fellow at George Mason University’s Mercatus Center.

In the U.S., Congress has held numerous hearings and briefings seeking to get a better understanding of AI and its potential uses, along with perspective on what the potential pitfalls might be. Predictions on the future of AI range from economically transformative to apocalyptic, leaving lawmakers in a challenging position to try to come with up the country’s first set of regulations for the industry to operate under.

Countries around the world have been trying to figure out how to handle AI amid rapid advancements and huge investments being made into the tech. The passage of the EU law will likely add pressure for other governments to act on AI despite its relative novelty as a booming industry.

“EU’s law is certain to put pressure on other countries, especially the U.S., to come up with their own regulations. The question is, what type of regulation? Big players are going to put pressure on the Congress to ‘compensate’ for the EU’s constraint in the name of ‘innovation,’” said Hamid Ekbia, director of the Autonomous Systems Policy Institute at Syracuse University and the leader of the Academic Alliance for AI Policy.

Lawmakers in both parties have expressed interest in and introduced their own legislation trying to take on a number of the potential problems, but there has not yet been anything that has come close to making it to President Joe Biden’s desk for passage.

Biden has taken his own steps to address artificial intelligence without Congress, signing an ambitious executive order last year establishing new standards for AI safety and security through reporting requirements, commitments to abiding by safety standards from private companies and labeling of AI-generated content.

Congress has been receptive to some of the input from the private sector and also been wary about balancing innovation and regulation to avoid giving up the United States’ place as one of the world leaders in AI’s development. Leaders of the industry have also been supportive of regulating AI, though they have also sought to ensure any new laws work in their favor.

“The balance between regulation and innovation is hard to accomplish. But it is wrong to conclude from this that we should maintain the status quo, allowing a few big players to hold their monopoly and self-regulate in the name of innovation,” Ekbia said.

Another factor complicating efforts to legislate artificial intelligence is how quickly it has changed and developed over the last several years, a trend that is unlikely to change moving forward as companies continue to invest huge sums of time and money into development.

Creating regulations is never a quick process in the U.S., where ideas and bills can take years and multiple separate efforts to make it through the legislative process and be signed into law. Even people directly involved in the industry have a hard time predicting where AI’s capabilities could be in five years, making it effectively impossible for lawmakers to predict every possible scenario and drawback of potential regulations.

“The name of the game right now is uncertainty,” Mittelsteadt said. “This is huge and with anything that’s this big, it’s going to be uneven. There’s gonna be highs, there’s gonna be lows.”

This post was originally published on this site

Similar Posts