Yazılar

EU Reviews Anthropic’s Mythos as Cybersecurity and Banking Risks Draw Scrutiny

/in /tarafından

The European Commission is actively evaluating Anthropic’s advanced AI model Mythos, signaling that European regulators are moving quickly to assess how next-generation cyber-capable artificial intelligence may affect financial stability, cybersecurity policy, and broader digital governance.

According to European Economic Commissioner Valdis Dombrovskis, Commission officials have already met with Anthropic to review the technical capabilities and potential policy implications of Mythos, an AI system reportedly designed to identify software vulnerabilities and code flaws at unprecedented speed.

Regulatory concern centers on the possibility that tools like Mythos could dramatically accelerate offensive cyber operations if misused, particularly against critical sectors such as banking, public infrastructure, and enterprise systems. Security analysts warn that highly capable vulnerability-discovery models may compress attack timelines, allowing malicious actors to identify and exploit weaknesses far faster than traditional defensive structures can respond.

Although Mythos has reportedly not yet been deployed within European banking institutions, the EU’s rapid engagement reflects a broader strategic priority: preventing AI-driven cybersecurity disruption before systemic exposure expands. The review is likely to intersect with the EU’s evolving AI Act, cyber resilience frameworks, and financial sector digital safeguards.

The situation highlights an emerging regulatory frontier where AI is no longer viewed solely as an economic or productivity tool, but also as a potential strategic cyber capability requiring oversight comparable to critical infrastructure technologies.

Europe’s response could become an important benchmark globally. If regulators conclude that advanced cyber-oriented AI systems require tighter deployment controls, transparency obligations, or sector-specific restrictions, Mythos may become one of the first major tests of how governments govern dual-use AI models.

Poland urges Brussels to probe TikTok over AI-generated content

/in /tarafından

Poland has asked the European Commission to investigate TikTok after the platform hosted artificial intelligence–generated content calling for Poland to leave the European Union, which authorities said was almost certainly Russian disinformation.

Polish officials said a TikTok profile featuring videos of young women dressed in Polish national colours and promoting an exit from the EU had gained traction in recent weeks before disappearing from the platform. Warsaw argues the content posed risks to public order, information security and democratic processes both in Poland and across the EU.

In a letter to the Commission, Deputy Digitalization Minister Dariusz Standerski said the use of synthetic audiovisual material and the way it was distributed suggested TikTok was failing to meet its obligations as a “Very Large Online Platform” under EU law. A Polish government spokesperson said the videos contained Russian linguistic patterns, pointing to a coordinated disinformation effort.

TikTok said it has been in contact with Polish authorities and removed content where it violated platform rules. A Commission spokesperson confirmed receipt of Poland’s request, noting that under the Digital Services Act (DSA), very large platforms must assess and mitigate risks linked to their services, including those arising from AI-generated content. The Commission added that it had already sought information from TikTok and other platforms in March 2024 on how they address AI-related risks.

EU governments have stepped up scrutiny of social media platforms amid concerns over foreign interference in elections and domestic politics. Last year, the Commission opened formal proceedings against TikTok, which is owned by ByteDance, over its handling of election-related risks, including during Romania’s 2024 presidential vote.

Poland is now urging Brussels to open new proceedings under the DSA, which requires major platforms such as TikTok, Meta’s Facebook and X to remove harmful content. Breaches can result in fines of up to 6% of a company’s global annual turnover.

EU Considers Pausing Parts of Landmark AI Act Amid Pressure from U.S. and Big Tech

/in /tarafından

The European Commission is considering pausing parts of its landmark Artificial Intelligence Act, following growing pressure from U.S. officials and major tech companies such as Meta and Alphabet, the Financial Times reported on Friday.

According to the report, the move comes after months of lobbying from Silicon Valley giants and warnings from the Trump administration that strict EU regulations could strain transatlantic trade relations.

A senior EU official told the FT that Brussels has been “engaging” with Washington on potential adjustments to the AI Act and related digital regulations as part of a broader simplification effort, which is expected to be adopted on November 19.

The AI Act, which became law in August 2024, is the world’s first comprehensive framework to regulate artificial intelligence technologies. It categorizes AI systems by risk level — from minimal to unacceptable — and imposes restrictions on areas like facial recognition, biometric surveillance, and generative AI transparency.

While a European Commission spokesperson had previously dismissed calls for delays, officials are now reportedly weighing temporary pauses for specific provisions, particularly those affecting companies developing large AI models.

An EU spokesperson told the FT that “various options” are being discussed but emphasized that the bloc remains “fully behind the AI Act and its objectives.”

The proposal reflects Europe’s balancing act between maintaining AI safety and innovation leadership while addressing geopolitical and trade pressures from the United States and industry stakeholders.