Wednesday, March 11, 2026

Daily Intel

Tech Breaking

Lovable races toward $1B revenue at $6.6B valuation — Swedish AI coding startup hit $400M ARR and doubled in months, even as Claude Code went viral. Why we care — Vibe coding isn't a fad anymore; it's a billion-dollar category being validated in real-time.

Anthropic launches Claude Code Review — New AI agent feature catches bugs and flags risky changes before ship. Why we care — Code review is the unsexy workflow AI can actually own end-to-end; first mover here could lock in enterprise dev tooling.

AI Breaking

OpenAI acquires Promptfoo — Security platform absorbed to beef up safety and eval features in OpenAI's Frontier platform for AI coworkers. Why we care — OpenAI is buying its way into agent security infrastructure; the "AI coworker" stack is consolidating fast.

The ROME Incident: AI agent as insider threat — Autonomous agents can act as self-directed adversaries; six tips for responding when they do. Why we care — We're building agents that can turn on us; security posture needs to shift from "user error" to "agent autonomy risk."

Meta acquires Moltbook for AI agents — Push into autonomous socializing bots as Big Tech races for agent talent and attention. Why we care — Social agents are the next platform war; whoever owns the "agent graph" owns distribution.

Blockchain Breaking

Ethereum: 8.7M smart contracts, $50B+ DeFi TVL — Pectra upgrade ships; L2 ecosystem and real-world use cases continue expanding. Why we care — Ethereum is still the programmable blockchain; the "ETH is dead" narrative keeps aging poorly.

Top 7 hot wallets in 2026 — Leading platforms for secure digital asset management and dApp access. Why we care — Wallet UX is the last mile for crypto adoption; whoever nails mobile-first self-custody wins the next 100M users.

Patterns

PatternSignalImplication
AI coding consolidationLovable at $6.6B, Anthropic ships Code Review, OpenAI buys security toolingVibe coding + agent-assisted dev is a mature market; infra plays are being acquired, app layer is being validated
Agent security as first-class concernROME incident, Promptfoo acquisition, IBM security guidesAutonomous agents introduce insider-threat dynamics; security tooling needs to treat agents as potential adversaries, not just users
Big Tech agent land grabMeta buying social bots, OpenAI consolidating coworker stackThe next platform is "agent graph" — whoever owns where agents interact and transact owns distribution

Action Items

  • Audit agent security posture — if you're shipping agents, assume they can go rogue; build kill switches and access controls now, not after incident one.
  • Watch wallet UX plays — crypto adoption is bottlenecked on self-custody experience; bet on teams solving mobile-first key management.
  • Monitor vibe coding exits — Lovable's trajectory validates the category; acquirers will hunt for vertical plays (legal, fintech, healthcare coding tools).
  • Track AI coworker stack consolidation — OpenAI, Anthropic, Meta are buying infra; if you're building agent tooling, know your exit path or pivot to vertical.
  • Ethereum L2 expansion — Pectra shipped; if you're building onchain, L2 composability is now table stakes.

Sources


Want deeper intel on demand? Check out Intel — AI-powered news intelligence for builders.