In an unprecedented show of industry unity, over 100 leading technology, cybersecurity, and financial organizations released a joint open letter titled "A call for collective action on cyber defense," warning that rapidly advancing artificial intelligence systems could soon dramatically escalate cyber threat levels globally. Organized by OpenAI and co-signed by major technology firms including Microsoft, Google, Anthropic, Amazon Web Services, IBM, Cloudflare, CrowdStrike, and Palo Alto Networks, as well as financial giants like Visa and Mastercard, the statement warns that defenders have only a brief window to prepare before offensive AI capabilities outpace existing defensive security measures.
The coalition emphasized that as AI models become more capable and accessible, malicious actors will gain cheaper and faster tools to scan, identify, and exploit vulnerabilities across interconnected networks. Critical infrastructure facilities—such as healthcare systems, municipal water treatment plants, power grids, and core internet hardware—face the most severe risks from automated hacking tools. Traditional cybersecurity practices, which already struggle against unpatched software and outdated system configurations, could easily be overwhelmed if organizations fail to modernize their security posture immediately.
To mitigate these emerging risks, the open letter outlines a multi-pronged action plan. It urges private enterprises and public sector institutions to treat cyber defense as an immediate strategic priority by applying essential security patches, implementing strict access controls, and adopting modern zero-trust architecture. Furthermore, the companies called on local, national, and international governments to establish trusted access programs that allow vetted security defenders early access to advanced AI models for defensive testing and mitigation. While AI introduces significant security risks, the coalition stressed that when deployed effectively by defenders, AI-driven tools can also help identify and resolve long-standing software vulnerabilities before cybercriminals can exploit them.
Meta has agreed to pay up to $17.1 billion to settle nationwide US lawsuits alleging that Facebook and Instagram were designed with addictive features harmful to youth mental health.
OpenAI announced that its upcoming artificial intelligence model, Astra, has become the first to reach the 'Critical' cybersecurity capability tier under its Preparedness Framework. The classification follows tests where Astra independently identified zero-day vulnerabilities and created functional exploit chains, leading to restricted access during deployment.
Semiconductor giant NVIDIA has entered an expanded agreement with MediaTek, backed by a $3.5 billion convertible bond investment, to integrate custom AI chips into NVLink-powered data center architecture.
OpenAI has announced plans to terminate its contract providing AI models to the popular coding platform Cursor after its acquisition by SpaceX. The company cited compliance and trust concerns regarding past contractual issues with Elon Musk's ventures.
Imaginary Image