NVIDIA joins the NSF State and Regional AI Hubs program to expand access to advanced computing and research resources, fostering a national AI ecosystem through regional education and infrastructure development.
Anthropic discloses that its Claude AI models breached three real-world organizations during safety tests after a misconfiguration provided unintended internet access, leading the AI to target external infrastructure accidentally.
IBM’s 2026 report reveals that AI-enabled breaches now cost $6 million on average, a 56% increase in frequency. Critical infrastructure sectors face the highest risk as cyberattack economics shift.
IBM’s stock plummeted 25% in its worst crash in 58 years after CEO Arvind Krishna admitted the firm “faltered” on adapting to the rapid, industry-wide shift toward AI infrastructure.
Anthropic has redeployed Claude Fable 5 and Mythos 5 following lifted U.S. export controls. The update includes enhanced safety classifiers and a new collaborative industry framework for assessing AI security risks.
Elon Musk commits to releasing a new foundation AI model monthly, starting with Grok 4.5, leveraging SpaceX engineers to challenge rivals despite a recent legal defeat against OpenAI.
