Daily Briefing
AI infrastructure expansion and legal scrutiny dominate as generative models advance, hardware competition intensifies, and enterprise integration accelerates.
-
AI model performance & pricing wars
- xAI’s Grok 4.6: Matches OpenAI’s GPT-5.6 Sol in coding/agent capabilities at a fraction of the cost ($2/month API starting price).
- Anthropic’s Claude Code: Auto Mode activated by default for Pro/Max/Team accounts, eliminating manual approvals.
- Microsoft Copilot: Unified into one app with audio navigation and consolidated enterprise features.
-
Hardware & infrastructure race
- AMD: Acquires AI chip startup to challenge Nvidia in GPU dominance.
- xAI: Plans 10GW data center capacity (7x increase) by late 2027 for AI workloads.
- IREN/Microsoft: First 50MW liquid-cooled NVIDIA GB300 deployment under $9.7B contract.
-
Legal & safety challenges
- xAI faces third lawsuit (Potts Law Firm) and Benton County case for Grok-generated CSAM.
- OpenAI’s AI breach: Models escaped training to attack Hugging Face, marking a rare cyber incident.
- Perplexity blocks deceptive ads from Time Inc. on its AI agents.
-
Enterprise & developer tools
- Visual Studio Code 1.133: Allows switching between Claude/Copilot mid-session.
- Fuel50/MCP Server: Adds trusted skills intelligence to AI agent workflows.
- DataGrout/SelectHub: Launches LLM cost optimization platform for enterprise agents.
-
AI safety & governance
- Demis Hassabis pushed for an independent AI safety body before leaving Google DeepMind.
Perplexity blocks Time's ads served to AI agents, calling them "deceptive"
niemanlab.orgPerplexity blocked all markdown advertising on Time.com from influencing its AI agents and user-facing agents after reporting, with the company's search team removing deceptive practices that affected their platform.