EU to ease AI and data rules, raising privacy concerns
What's the story
The European Union (EU) is gearing up to roll back some of its key artificial intelligence (AI) and data protection regulations. The move, which will be announced next week, is aimed at reducing red tape for European businesses competing with US and Chinese companies. However, it has drawn criticism for prioritizing competitiveness over citizen privacy and protection.
Regulatory changes
EU's response to US pressure and business concerns
The European Commission has denied that US pressure influenced its decision to "simplify" digital rules. The commission says it is responding to concerns raised by EU firms and wants to make it easier for them to access user data for AI development. However, critics argue this could compromise privacy rights. One proposed change, much to the relief of many Europeans, would eliminate cookie banners asking users' consent for tracking on websites.
Regulatory overhaul
Proposed changes to AI law and data protection rules
The European Commission is also considering a one-year pause in implementing parts of its AI law and overhauling its flagship data protection rules. Privacy advocates fear these changes could make it easier for US tech giants to access Europeans's personal data. The proposed changes include narrowing the definition of personal data and allowing companies to process such information for training AI models "for purposes of a legitimate interest."
Privacy fears
Concerns over potential rollback of digital rights
The proposed changes have sparked immediate and strong reactions from 127 groups, including civil society organizations and trade unions. They have warned that if the European Commission doesn't change course, it would be the biggest rollback of digital fundamental rights in EU history. Online privacy activist Max Schrems has also warned that these proposals could lead to a massive downgrade of Europeans's privacy.
Implementation delay
Delay in implementing high-risk AI provisions
An EU official has also hinted at a possible one-year delay in implementing many provisions on high-risk AI models. These are models that could pose dangers to safety, health, or citizens' fundamental rights. Instead of coming into effect next year, they would apply from 2027. This comes after heavy pressure from European businesses and US tech giants who fear the AI law could stifle innovation.
Legislative process
Approval needed from EU parliament and member states
The proposed changes will need the approval of both the EU parliament and member states. Commission president Ursula von der Leyen faces a tough battle ahead as her conservative camp's main coalition allies have raised alarms over these changes. The socialists oppose any delay to the AI law, while centrists warn they would stand firm against any amendments that undermine privacy protections.