Global Courant 2023-05-13 00:07:42
Authorities around the world are racing to draft rules for Artificial Intelligence (AI), including the European Union, where the draft law on the subject faced a crucial moment on Thursday.
A European Parliament committee voted to strengthen the main legislative proposal, which is heading for approval, part of a multi-year effort by Brussels to create safeguards for Artificial Intelligence.
These efforts have accelerated as a result of rapid advances in chatbots, such as ChatGPT, that highlight the benefits and risks that the emerging technology can bring.
We take a look at the EU Artificial Intelligence Act:
The Artificial Intelligence Act, first proposed in 2021, will regulate any product or service that uses an AI system.
The act will classify AI systems according to four levels of risk, from minimal to unacceptable. More risky applications will face tougher requirements, including being more transparent and using accurate data.
Think of it as a “risk management system for AI,” said Johann Laux, an expert at the Oxford Internet Institute.
One of the main goals of the EU is to protect against any AI threats to health and safety, and to protect fundamental rights and values.
This means that some uses of AI are an absolute NO, such as “social profiling” systems that judge people based on their behavior or systems that aim to manipulate like interactive talking toys that encourage risky behavior .
The basic proposal was strengthened by voting to ban predictive police tools, which crunch data to predict where crimes will occur and who will commit them.
They also voted to ban remote facial recognition, with some law enforcement exceptions, such as preventing a specific terrorist threat.
The technology scans passers-by and uses AI to match their faces to a database.
The goal is to “avoid a controlled society based on AI,” Brando Benifei, the Italian lawmaker who helps lead the European Parliament’s AI efforts, told reporters on Wednesday.
“We consider the risks to be very high.”
Artificial Intelligence systems used in high-risk categories such as employment and education, which would affect the course of a person’s life, face difficult requirements such as being transparent with users and establishing risk assessment and measures. emollient.
The EU’s executive arm says most AI systems, such as video games or spam filters, fall into the low or no risk category.
The original 108-page proposal barely mentioned chatbots and simply requires them to be labeled so users know they’re interacting with a machine.
Provisions were later added to cover general-purpose Artificial Intelligence such as ChatGPT, subject to some of the same requirements as high-risk systems.
An additional provision is a requirement to fully document any copyright material used to teach AI systems how to generate text, images, video or music that resembles human work.
This would allow content creators to know whether their blog posts, e-books, scientific articles or pop songs have been used to train algorithms that power systems like ChatGPT. Then they can decide if their work has been copied and request correction.
The European Union is not a major player in recent AI development. This role is taken by the USA and China. But Brussels often plays a trend-setting role with regulations that tend to become de facto global standards.
By establishing common rules for Artificial Intelligence, Brussels is also trying to develop the market by instilling confidence in users, Laux said.
“If you can get people to believe in AI and the applications, they will also use it more,” he said, adding: “And, when they use it more, they will unlock the economic and social potential of AI “.
Violations will result in fines of up to 30 million euros ($33 million) or 6 percent of a company’s annual global revenue, which in the case of tech companies such as Google and Microsoft can run into the billions.
It could be years before the rules take full effect.
European Union lawmakers must now vote on the bill in a plenary session in mid-June. It then moves to tripartite negotiations involving the bloc’s 27 member states, Parliament and the Executive Committee, where it could face more changes as the parties clash over details.
Final approval is expected by the end of the year, or early 2024 at the latest, following a period for companies and organizations to adjust to the new legislation, which is often around two years.
Global Courantl
