TOTAL: {[ getCartTotalCost() | currencyFilter ]} Update cart for total shopping_basket Checkout

The Privacy Advisor | European Parliament vote pushes AI Act significant step forward Related reading: As generative AI grows in popularity, privacy regulators chime in



The EU took a significant step in its efforts to pass the world's first regulation of artificial intelligence. In a full plenary vote, the European Parliament agreed upon a negotiating position on the proposed AI Act, the final step before a three-way negotiating process — known as a trilogue — takes place.

With 499 votes in favor, only 28 against and 93 abstentions, the European Parliament settled on a text it will use to negotiate with the EU Council of Ministers and the European Commission in the trilogue process, the first of which takes place Wednesday evening.

At a press conference after the vote, European Parliament President Roberta Metsola said lawmakers worked tirelessly on "this first set of regulations to manage risk and develop lawful use of AI. All of this is perfectly consistent with our will to be world leaders in digital innovation based on EU values such as privacy and respect for fundamental rights."

"Europe is leading and will continue to lead in AI legislation," she added.

The AI Act takes a risk-based, tiered approach to regulating AI and includes outright bans of high-risk AI applications. Parliament's text prohibits the use of real-time biometric identification systems — a point of contention, particularly during the last week, among the European People's Party during Parliament's amendment process.

AI Act bans

Other bans include biometric categorization systems that use sensitive attributes, such as gender, race, ethnicity, religion and political affiliation. Predictive policing technology that uses profiling and geolocation, for example, would be banned, as well as so-called emotion recognition systems in the workplace, schools and law enforcement. Massive data scraping of facial images from the internet or closed-circuit television for facial recognition databases would also be prohibited.

Italian MEP Brando Benifei said, "All eyes are on us today," adding, "We want AI's positive potential for creativity and productivity to be harnessed but we will also fight to protect our position and counter dangers to our democracies and freedoms during the negotiations with council."

High-risk and generative AI

For high-risk AI, Parliament's text includes AI systems "that pose significant harm to people's health, safety, fundamental rights or the environment." The text also adds systems that could be used to influence voters and "recommender systems" used by social media companies with more than 45 million users.

Parliament's negotiating text would require providers of foundation models to assess and mitigate potential risks and register their systems in an EU database prior to release in the EU market. Generative AI systems, including the popular ChatGPT, would be required to disclose that content was AI-generated and help users distinguish between fake and real information.

To help support innovation, the text provides exemptions for research and promotes the use of regulatory sandboxes.

EU citizens would be granted the right to file complaints about AI systems, and covered organizations would be required to provide "explanations of decisions based on high-risk AI systems that significantly impact their fundamental rights."

The EU AI Office was reformed in Parliament's text and would monitor implementation of the AI regulation, according to a European Parliament news release.

Balance, convergence and inclusion

Co-rapporteur Dragos Tudorache said, "The signal we are sending is a very important one."

He highlighted three factors that undergirded Parliament's approach to the AI Act, including the balance between protecting EU citizens' fundamental rights — through redress, transparency and required impact assessments for high-risk applications — while promoting technological innovation through the use of regulatory sandboxes, a bottom-up approach to standard setting and governance.

"We have a very clear mandate for standard setting, something that in previous legislations of the EU such as (the General Data Protection Regulation), we didn't do and then we left companies to struggle with legal definitions," Tudorache said.

He said the definition of AI in Parliament's text aligns with definitions laid out by the Organisation for Economic Co-operation and Development and the U.S.

Second, Tudorache highlighted global convergence, noting that the "Brussels effect," where EU policy influences regulation and business practices around the world, "will not suffice." He said the EU needs to look at this differently, as AI technology is the same everywhere, and that like-minded democracies need to work together.

He also said AI must not leave citizens behind, that education on how to integrate AI into the workplace will be necessary, otherwise there will "be forces in society who will resist."

Critics react

Consumer advocacy organization Access Now characterized Wednesday's vote as "historic," and applauded the text's full ban of biometric surveillance and discriminatory AI, obligations to conduct fundamental rights impact assessments, requiring transparency, among others.

However, the organization has concerns about what it calls a "loophole" on Article 6 "that allows a self-assessment based carve out from high risk classification that undermines the already flawed risk-based approach."

Access Now EU Policy Analyst Caterina Rodelli said, "In this historic AI Act vote, the European Parliament called for a society free from mass surveillance. However, it has drawn a thick line between the haves and the have nots — the lack of bans for AI systems used in migration confirms the EU does not seek to protect fundamental rights when migrant people are the rights-holders."

What's next

The trilogue process starts Wednesday evening, but according to Euractiv's Luca Bertuzzi, "negotiations will intensify once Spain takes over the rotating presidency of the Council in July, as Madrid has made finishing up the AI law its top digital priority."

In a press conference Wednesday, co-rapporteur Tudorache said, "odds are very good we will finish negotiations by the end of the year," though he added companies will need time to implement the requirements and member states will need time "to set up their roles as market regulators."

Credits: 1

Submit for CPEs


If you want to comment on this post, you need to login.