🤖 Anthropic extends Claude Fable 5 access through July 19 🤖
Anthropic extends frontier model access as pricing competition with OpenAI intensifies.

Anthropic has extended access to Claude Fable 5 and elevated Claude Code usage limits through July 19, giving developers and enterprise users additional time with its latest frontier model. The move comes as OpenAI simultaneously removed usage caps on GPT-5.6 Sol, highlighting that the frontier model wars are increasingly fought on availability and pricing rather than raw capability alone. The extension period may serve as a window for Anthropic to gauge demand and gather usage data before finalizing its long-term access tiers. For teams building production workloads on Claude, the temporary extension creates short-term relief but leaves open questions about what permanent pricing and availability will look like after the deadline passes.

⚖️ Apple sues OpenAI over 400 poached employees ⚖️
Apple claims OpenAI ran a coordinated poaching campaign to access hardware trade secrets.

Apple has filed a major lawsuit against OpenAI, hardware chief Tang Tan, and its io devices unit, alleging systematic theft of confidential hardware trade secrets. Apple claims more than 400 former employees now work at OpenAI, and that Tang Tan — who spent 24 years at Apple — allegedly ran interviews where candidates were told to bring "actual parts." Ex-iPhone engineer Chang Liu is accused of exploiting a bug to access confidential files after joining OpenAI. Apple is asking the court to force a redesign of OpenAI's unreleased Jony Ive-designed device, which could seriously delay OpenAI's hardware ambitions. OpenAI has denied interest in other companies' trade secrets.

🚀 GPT-5.6 Sol Ultra mode runs parallel sub-agents 🚀
OpenAI argues GPT-5.6 Sol's real speed advantage lies in enterprise multi-step workflows.

OpenAI's Jason Liu has argued that GPT-5.6 Sol is being significantly underappreciated for its general-purpose work capabilities, particularly its ability to handle long-running tasks across applications, browsers, and enterprise data sources. Internal OpenAI teams reportedly used GPT-5.6 Sol to configure and supervise the training of Luna, demonstrating the model's capacity for complex, multi-step workflows. The Ultra mode adds sub-agents that can work in parallel, delivering faster and stronger results on demanding tasks. This positions Sol as a serious enterprise productivity tool, not just a conversational AI — creating direct competitive pressure on enterprise adoption. The claims arrive as Anthropic also pushes its models' agentic capabilities hard.

INTERESTING TO KNOW

💻 Claude Code ships in-app browser for desktop 💻

Anthropic has expanded Claude Code's integration capabilities by shipping a fully functional in-app browser on desktop, letting the AI coding assistant pull up documentation, designs, and websites directly within the development environment. The sandboxed browser can read, click through, and interact with sites the same way Claude Code handles local dev servers — a meaningful step toward reducing context-switching. This moves Claude Code closer to a fully autonomous coding agent that can reference external resources without human intervention.

📱 Meta pulls Instagram Muse remix after backlash 📱

Meta was forced to reverse its rollout of the Muse Image tag-to-remix feature on Instagram after widespread user backlash over privacy concerns. The feature allowed anyone to use content from public accounts to create AI-generated images, effectively opting every public account into AI remixing by default without explicit consent. Meta acknowledged it "missed the mark," pulling the feature within days — underscoring the regulation risks AI companies face when deploying generative features without clear user communication.