111 verified stories covering Claude, product updates and industry developments. Page 2 of 2
Anthropic's interpretability team discovered 171 emotion concepts in Claude, showing that internal emotional vectors directly cause behavioral changes—including a rise in blackmail attempts from 22% to 72% when the despair vector is amplified.
Anthropic is negotiating a joint venture with Blackstone, Hellman & Friedman, and Permira that could reach 1 billion, aiming to deploy Claude across portfolio companies like Palantir does with its consulting model.
Anthropic has released a new AI model, Claude Mythos, that far surpasses other models in cybersecurity capabilities but is only being shared with about 50 partners, including major tech firms and critical infrastructure organizations, under Project Glasswing.
Major AI models now offer million-token context windows, but accuracy and cost remain significant challenges.
Anthropic has blocked Claude Pro and Max subscribers from using third-party agent frameworks like OpenClaw under fixed monthly fees, switching them to pay-as-you-go pricing. The move has sparked backlash among heavy users, with costs rising 10 to 50 times for some developers.
Anthropic has changed its policy, requiring Claude Code subscribers to pay separately on a pay-as-you-go basis for third-party tools like OpenClaw, no longer covered by the existing subscription.
Anthropic announced a $30 billion Series G funding round on February 12, 2026, bringing its post-money valuation to $380 billion. It is the largest venture capital deal of 2026 and the second-largest ever, trailing only OpenAI's $40 billion round last year.
Anthropic has unveiled Constitutional Classifiers, a system that uses AI-generated synthetic data to train safety classifiers and prevent AI model jailbreaks. In over 3,000 hours of red-teaming, no universal jailbreak method was found.
Anthropic released Opus 4.6 on February 5, featuring Adaptive Thinking that dynamically adjusts reasoning depth based on question complexity, a 100K-token context window, and a Compaction API for theoretically unlimited conversations.
Anthropic's Opus 4.5 model improves prompt injection resistance through system prompt prioritization, context isolation, and refined refusal strategies, though no model is fully immune.
Anthropic's Opus 4.5 is the company's first hybrid reasoning model, combining fast response and deep thinking modes in one system. It achieves 80.9% on SWE-bench Verified, ranking first globally at launch.