Enterprise AI Gets Serious

The focus on governance and control in corporate AI is intensifying. SAP argues that proper enterprise AI governance, replacing statistical guesses with deterministic control, is key to securing profit margins (source). Microsoft is applying a similar principle of structured control with a new AI agent specifically for legal teams inside Word, designed to follow legal workflows rather than general commands (source). Meanwhile, the Pentagon has secured classified AI deals with major players like OpenAI, Google, and Nvidia, but notably left out Anthropic from this round (source).

Market Moves & Monetization

The AI market is showing signs of maturation and fierce competition. GitHub Copilot is shifting to a per-token charging model from a flat-rate subscription, signaling a move towards more granular, usage-based pricing for developer tools (source). In the legal tech arena, startup Legora has hit a $5.6B valuation, heating up its battle with rival Harvey (source). On the funding front, Anthropic is reportedly moving swiftly towards a potential valuation round that could exceed $900B (source), while Apple admits it was surprised by AI-driven demand for its Macs (source). Stripe is facilitating agent economies with Link, a digital wallet that autonomous AI agents can also use for spending (source).

Security, Scrutiny, and Slop

Security concerns are paramount. OpenAI announced new advanced security for ChatGPT accounts, including a partnership with hardware security key provider Yubico (source). Following its criticism of Anthropic for limiting access to its model Mythos, OpenAI itself is restricting initial access to its cybersecurity testing tool, GPT-5.5 Cyber (source). A severe new Linux threat, “CopyFail,” is also causing global concern (source). On the ethics front, a report details how Christian content creators are outsourcing AI-generated “slop” to gig workers on platforms like Fiverr (source).

Musk vs. OpenAI: The Trial Continues

The high-profile lawsuit between Elon Musk and OpenAI dominated headlines. Musk testified that his company, xAI, trained its Grok model on OpenAI models, bringing the concept of “distillation” into the courtroom (source). Legal analysts suggest Musk is unlikely to win his case (source), and a major procedural error by Musk’s team while the jury was out of the room may have compounded their problems (source).

In Other News…

• ChatGPT Images 2.0 is finding particular success in India for creating personal visuals (source).
• Google’s Gemini AI assistant is expanding into millions of vehicles (source).
• MIT startup Beacon Biosignals is using AI to map the brain during sleep for disease diagnosis (source).
• MIT researcher Olivia Honeycutt is investigating how language shapes our understanding of the world (source).
• BioticsAI’s founder discussed the realities of building an AI company in the heavily regulated healthcare space (source).

Editorial Take: Today’s stories paint a picture of AI moving from a wild west of innovation into a phase of consolidation, regulation, and commercial pragmatism. Enterprises and governments are demanding control and security (SAP, Microsoft, Pentagon), while the market is rationalizing through new pricing models (GitHub Copilot) and sky-high valuations for specialized players (Legora, Anthropic). The courtroom drama between Musk and OpenAI underscores the intense competition and contentious IP battles defining this era. The overarching theme is a shift from what AI can do to how it should be managed, paid for, and deployed responsibly.