Everything the newsroom published, in chronological order. Each item carries origin, sources and reading time.
RESEARCH Microsoft's Agensh reaches 55% test-pass rate with 1,024 agents
RESEARCH SWE-Serve benchmark reveals 23-point gap between local tests and production serving
RESEARCH A2M attack hijacks MCP agents with poisoned tool metadata at 93.6% success
RESEARCH Harness choice drives 5x cost gap while coding agent success rates stay flat
INDUSTRY GitLab Duo adds self-hosted models through Microsoft Foundry
RESEARCH Flash-dLLM cuts diffusion LLM inference latency 11× with fused kernels
INDUSTRY Linear halved CI runner time despite test suite quadrupling
INDUSTRY Cloudflare Worker Previews isolate agent changes in production-like branches
INDUSTRY Databricks Genie One MCP reaches general availability
FUNDING Prevalent AI $22M: data fabric for agentic AI as 40% of agent projects face cancellation
MARKET Amazon completes $50B OpenAI investment; contingent milestone unlocks signal IPO track
BREAKING Anthropic in talks with Microsoft to run Claude on custom Maia 200 chips via Azure
MARKET Nvidia raises GPU server prices 15%+ for largest customers, citing memory cost surge
INDUSTRY GLM-5.3 Beats Sol on Cost; Cascade Reaches 85.9% at $6.61
MARKET AI Data Center Bonds Hit Junk-Level Yields as Funding Costs Surge
FUNDING Castelion hits $13B on $1B Series C for hypersonic missile production
FUNDING Anthropic Eyes October IPO at $2T Valuation; Q2 Hits Operating Profit
POLICY Texas Orders Statewide Data Center Audit; 27 States Push Back on AI Infrastructure Costs
CHIPS Marvell partners with Google on TPU ecosystem silicon; $12.2B warrant through 2033
FUNDING Broadcom in talks for $60B+ SPV debt deal for AI chip financing to Anthropic, others
MARKET Broadcom seeks $60B+ debt to finance AI chips for Anthropic, signaling scaling constraints
MARKET AI stocks face data center regulation headwinds as yields rise; GE Vernova, infrastructure names pressured
MARKET Chinese AI models DeepSeek, Qwen close US performance gap at 90% lower cost
CHIPS Desktop CPU shipments crater 20% YoY; AMD hits record 34.9% share despite market collapse
MARKET Chinese AI models cut token costs 5–12x below Claude/GPT; developers shifting defaults by August 2026
FUNDING Wispr Flow raises $280M at $2B; Canto speech model drops error rates to 5-10% in noise
INDUSTRY DoorDash Cuts Safety Incidents 50% With Cascade Moderation
BREAKING CISA: Ray CVE-2025-62593 (RCE via browser) actively exploited; federal patch deadline Aug 20
MARKET Google commits $12.2B to Marvell for custom AI chips through 2033; $120B revenue potential
FUNDING Muon Space raises $250M Series C at $1.5B; 100% launch success, 500 sat/year capacity by 2027
BREAKING Spline V2 integrates MCP; Claude Code can now edit 3D scenes live in real-time with full undo/sync
FUNDING Rillet raises $100M Series C at $1B; doubles ARR in 3 months, replaces legacy ERPs with agentic finance
RESEARCH GLM-5.3 + GPT-5.6 Sol cascade: 85.9% solve rate, 21% cheaper than Sol alone
RESEARCH GLM-5.3 vs Claude Fable 5 on DeepSWE: 69% accuracy at 5.4x lower cost; routing strategy wins
FUNDING Emesent closes $17M ($10M equity + $7M NRFC venture debt) for Cortex AI and Aura; GPS-denied autonomy deployed 200+ mine sites globally
BREAKING Anthropic deploys Claude Mythos 5 in Claude Security; $35M open-source vulnerability fund, enterprise access expands
RESEARCH LangSmith Trace Judge via Fireworks: Qwen Fine-Tune Hits Frontier Performance at 100x Lower Cost
FUNDING NVIDIA pays $6B to license Poolside's AI model factory; founders stay to run infrastructure play INDUSTRY LangChain cuts evaluation cost 100x with a fine-tuned open model