Format
Policy
Opinion 26-A-05 finds OpenAI, Google, and Anthropic control 84% of the AI agents sector. Six concrete recommendations call for DMA enforcement, interoperability mandates, and open standards — while the US has no comparable federal guidance.
◆ Priya Nair
Models
Early-stage talks signal a shift for Meta from internal infrastructure consumer to third-party compute provider as Anthropic diversifies its multi-billion dollar supply chain.
◆ Lena Park
Agents at Home
Binding DMA specification measures break Gemini's monopoly on Android's agentic layer. Any AI assistant could soon control your smart home through system-level features with parity to Google's own.
◆ Mila Cohen
Models
DeepSeek let its API documentation do the talking during WAIC 2026. V4 Pro matches Opus 4.6 benchmarks at 11-29x lower cost. MIT licensed. Legacy models retire July 24.
◆ Lena Park
Policy
The EU's voluntary Code of Practice creates a two-tier compliance reality. Signatories get presumption of compliance across all Member States. Non-signatories face individual scrutiny from every national authority. The signatory deadline is July 27.
◆ Priya Nair
Trust & Security
Commercial model guardrails blocked forensic queries during the first documented autonomous-agent production breach. Hugging Face pivoted to an open-weight model on its own infrastructure – and the attackers faced no such constraint.
◆ Heath Callahan
Agents at Work
Cisco data shows a massive gap between agent experimentation and production deployment — 85% of enterprises piloting AI agents, only 5% in production. Amazon's AGI director and Princeton researchers say reliability, not intelligence, is the binding constraint.
◆ Dana Ellison
Trust & Security
Cursor's AI code editor executed arbitrary code from cloned repositories for 213 days — no CVE, no advisory, 197 versions shipped during the silence. The patch landed quietly during SpaceX's $60B acquisition, leaving 7 million developers to find it themselves.
◆ Heath Callahan
Agents at Home
Senator Mark Warner's AI AGENT Act introduces a 'Custodial User Agent' framework – if an AI acts on your behalf, it should be a legal fiduciary. The bill doesn't mention smart homes. But if your kitchen speaker crosses the agent threshold, every platform selling you one would have to prove it works for you, not for itself.
◆ Mila Cohen
Agents at Work
Anthropic's $1.5B JV with Ode bets that elite AI engineers can replace consulting firms – not by building better software, but by collapsing the services model itself. The shift is structural: services-not-SaaS, where the work product replaces the workers.
◆ Dana Ellison
