The European Union is willing to consider ways for artificial intelligence companies to coordinate on safety without violating competition rules, as policymakers grapple with calls from leading developers for greater cooperation over the risks posed by increasingly powerful AI systems.
Teresa Ribera, the European Commission executive vice president responsible for competition policy, indicated that Brussels is open to discussing the issue, according to POLITICO Europe. Any arrangement would have to reconcile cooperation on safety with the EU’s longstanding objective of preserving competitive markets and preventing excessive concentration of economic power.
The debate has intensified as executives at major AI developers have called for closer coordination over the development and deployment of advanced models. Such cooperation could involve safety standards, testing procedures or agreements intended to reduce the risk of releasing systems whose capabilities are not sufficiently understood.
Competition law, however, generally limits coordination among rivals when it could restrict competition. That creates a potential legal complication for AI companies seeking to collaborate on issues that could also influence how quickly models reach the market or what requirements companies must satisfy before releasing them.
Ribera has argued separately that governments should take warnings from the AI industry seriously. At an event in Madrid this week, she called recent signals from major technology companies about the potential loss of control over advanced AI systems “absolutely worrying,” according to reporting on her remarks. She advocated transparent and secure governance while emphasizing the need to avoid concentrating power in a small number of companies.
The European Commission’s approach is significant because Ribera oversees the EU’s competition portfolio, including efforts to modernize competition policy and strengthen enforcement against anticompetitive practices.
The discussion in Europe comes as a similar debate unfolds in the US. Anthropic Chief Executive Officer Dario Amodei has called for a narrowly tailored antitrust exemption that would permit AI developers to coordinate on certain safety issues, according to separate reporting by Reuters.
We’d love to be your preferred source for news.
Please add us to your preferred sources list so our news, data and interviews show up in your feed. Thanks!
US regulators have shown more skepticism. Federal Trade Commission Chairman Andrew Ferguson said Tuesday that requests from major AI companies for both regulation and antitrust protection warranted scrutiny, arguing that such measures could potentially shield established companies from new competitors. Ferguson said he was expressing his personal view.
OpenAI, meanwhile, has taken a different position. Chris Lehane, the company’s global policy chief, said OpenAI has been working with Anthropic and Google DeepMind on AI safety and doesn’t believe an antitrust waiver is necessary for those discussions, according to Bloomberg News reporting published this week.
Read more: FTC Chair Casts Doubt on Antitrust Exemptions for AI Industry
The competing positions underscore a broader problem facing regulators: cooperation among developers may make it easier to establish common safeguards, but agreements governing model testing, capability thresholds or the pace of deployment could also affect competition and make it harder for smaller companies to challenge established players.
Antitrust specialists have noted that competition law can permit certain forms of cooperation when they generate legitimate benefits, though arrangements among competitors can attract scrutiny when they influence market access or constrain independent business decisions. The issue becomes particularly sensitive in AI because safety standards themselves can affect development costs and determine when products can be released.
For the EU, that tension fits into a broader effort to regulate AI while maintaining competition in a market increasingly dominated by companies with enormous computing resources and capital requirements. Ribera has repeatedly emphasized the risks associated with concentrated digital power while arguing that technological development should ultimately benefit society more broadly.
According to POLITICO Europe’s reporting, Brussels’ willingness to discuss an antitrust solution does not amount to an endorsement of a specific exemption. Rather, it signals that European officials are prepared to examine whether competition rules need to accommodate narrowly defined cooperation intended to address AI safety.
Source: POLITICO
Add as Preferred Source