Meta and Anthropic AI Updates

The Pentagon's blacklisting of Anthropic stems from the company's refusal to remove ethical safeguards from its AI models. Anthropic insisted on "hard limits" to prevent use for autonomous lethal weapons and mass domestic surveillance, which conflicted with the Defense Department's demand for unrestricted "any lawful use" access. The standoff led to Defense Secretary Pete Hegseth branding the San Francisco-based company a "supply chain risk," a designation previously unused for American companies. Anthropic's explosive revenue growth saw its annualized run rate jump from $9 billion at the end of 2025 to over $19 billion by March 2026. This surge is largely credited to strong enterprise adoption of its AI models and its agentic coding tool, Claude Code. The growth occurred despite the high-profile dispute with the Pentagon, with the company's main app even hitting the top of Apple's download charts. The AI model race is defined by the close competition between GPT-5.4, released March 5, 2026, and Gemini 3.1 Pro, released February 19, 2026. Benchmarks show a split decision: GPT-5.4 is the first model to surpass human performance on desktop computer tasks, while Google's Gemini 3.1 Pro leads in abstract reasoning and costs significantly less for pro-tier usage. Apple's redesigned Siri is a core component of "Apple Intelligence," its new suite of AI features. The update allows for more natural, context-aware conversations and enables Siri to perform actions within and across apps, like editing a photo and sending it to a contact. For more complex queries, Siri can now tap into ChatGPT's knowledge base. Meta's Llama 3 family represents a major push into multilingual capabilities, with its training data comprising over 15 trillion tokens, including high-quality data from over 30 languages. The architecture uses a vocabulary of 128,000 tokens to improve language encoding efficiency and performance in non-English languages. While Amazon's AWS CEO has stated AI should amplify, not replace, employees, the company has cut roughly 30,000 corporate roles since late 2025 amid an AI-driven efficiency push. This includes at least 100 white-collar positions within its robotics division and the halting of its "Blue Jay" multi-arm robot project.