OpenAI has widened access to GPT-5.4-Cyber, a cybersecurity-focused variant of its flagship model that is being offered only to vetted defenders. The rollout sharpens the company’s enterprise security push as AI vendors compete to build tools for vulnerability research, triage and incident response.
Amazon has agreed to invest another $5 billion in Anthropic, with the potential for up to $25 billion more tied to performance milestones, according to Reuters. The deal deepens a relationship that already links one of the world’s largest cloud providers with one of the fastest-growing AI labs, and it arrives as the race to secure training and inference capacity becomes a defining constraint in artificial intelligence.
European lawmakers voted on March 26, 2026 to delay parts of the AI Act, including some high-risk system requirements and watermarking deadlines, creating more breathing room for AI companies even as the bloc’s flagship AI rulebook continues to take shape.
The latest clash over AI regulation is playing out in Utah, where a Republican lawmaker is still pushing child-safety rules even after the Trump administration helped stall his proposal. The fight highlights a larger question now before Congress and the states: whether Washington will set a federal framework or move to preempt local AI laws altogether.
OpenAI is widening access to GPT-5.4-Cyber, a cyber-permissive version of its flagship reasoning model, as it scales its Trusted Access for Cyber program to more verified defenders. The change gives security teams a more capable model for legitimate reverse engineering and vulnerability analysis while keeping the rollout gated by identity and trust checks.
OpenAI has unveiled a major India initiative that pairs local AI infrastructure with enterprise and education deals, including a plan to start with 100 megawatts of data center capacity and open offices in Mumbai and Bengaluru later this year.
EU negotiators are discussing a compromise that would simplify how the AI Act interacts with sector-specific rules, a sign that Brussels is recalibrating the bloc’s flagship AI law under growing pressure from industry and member states.
Microsoft has unveiled a $10 billion investment in Japan through 2029, pairing new AI infrastructure with cybersecurity and workforce programs in one of its largest regional commitments to date.
OpenAI is rolling out GPT-5.4 mini in ChatGPT, making the model available to Free and Go users through the Thinking option and using it as a fallback for higher-tier users when GPT-5.4 Thinking reaches rate limits.
OpenAI said Japanese internet company CyberAgent is using ChatGPT Enterprise across nearly all departments, with monthly active usage reaching 93% as the company integrates generative AI into research, drafting, code review, and product development.
