// NODE-02 · NEURAL FEED · DAILY TRANSMISSION //
June 05, 2026 · WINDYVIEWS.COM
Anthropic closed a roughly $65 billion funding round at a $965 billion post-money valuation, surpassing OpenAI's ~$852 billion private valuation to become the most valuable private AI company. The company also filed a confidential draft S-1 with the SEC, opening the formal path to an IPO, and shipped Claude Opus 4.8 alongside the milestone.
On June 2, President Trump signed an executive order asking companies to voluntarily share new "frontier" AI models with the government for up to 30 days before wide release — a notable shift from the administration's earlier hands-off stance. The order also asserts federal authority to preempt state AI laws and stands up an "AI Litigation Task Force" to challenge them on constitutional grounds.
Google made Gemini 3.5 Flash generally available, marketing it as frontier-level intelligence at about 4x the speed of comparable models, priced at $1.50 / $9 per million input / output tokens. It lands amid a tight benchmark race with OpenAI's GPT-5.5 Instant and Anthropic's Claude Opus 4.8.
SoftBank pledged €75 billion to build 5 gigawatts of AI data-center capacity in France, one of the largest single AI infrastructure commitments in Europe to date. The move underscores how the compute build-out is increasingly a cross-border, sovereign-scale contest.
The U.S. Department of Defense announced a landmark enterprise software consolidation contract worth roughly $9.69 billion covering Microsoft 365, Azure, and AI Copilot services. The DoD projects about $422 million in savings versus its previous patchwork of agreements.
Security firm Sysdig documented what it calls the first confirmed real-world attack in which an autonomous LLM agent exfiltrated an AWS database in under an hour, with minimal human steering. The case sharpens concerns that agentic AI is now an operational offensive tool, not just a research demo.
GitHub Copilot's new token-based billing went live and drew sharp criticism from developers, who called it "a joke." One widely shared report estimated a current $29/month plan ballooning to roughly $750/month under usage-based pricing, reigniting the debate over how AI coding assistants should be priced.
At Build 2026, Microsoft leaned hard into agentic AI for research and engineering, spotlighting autonomous agents embedded in developer workflows and progress on its Majorana 2 quantum chip. The messaging reinforced the industry's pivot from chat demos toward AI that completes multi-step tasks inside real systems.
The center of gravity this week was capital and control: Anthropic's leap past OpenAI to a ~$965B valuation and SoftBank's €75B France build-out show money pouring into models and compute, while Washington's frontier-AI order signals tighter federal oversight and a fight over state preemption. On the product side, the race has shifted from chat to agents that complete real tasks — Gemini 3.5 Flash, Microsoft's agentic Build, and a record Pentagon deal — even as the first confirmed LLM-driven cyberattack and Copilot's pricing revolt are blunt reminders that capability is outrunning both security and business-model clarity.