Washington Draws the AI Regulatory Map

Alibaba and Tencent lost $66 billion in 24 hours because investors couldn't get a straight answer on when AI spending turns into revenue. Meanwhile, US VCs raised $34B+ in new mega-funds and poured money into cybersecurity at a pace that suggested they've stopped asking that question entirely. The White House handed them a regulatory gift: a framework to kill state AI oversight before it gets expensive. These three stories are connected. They're all about who gets to define the rules of the AI economy — and who pays when the bill comes due.

 Alibaba and Tencent Lost $66 Billion in 24 Hours

China's two most valuable tech companies shed $66 billion in combined market cap in roughly 24 hours after back-to-back earnings calls failed to answer the only question investors care about: when does AI spending translate into revenue? Tencent dropped $43 billion in a single day — its worst session in nearly a year. Alibaba fell another $23 billion, its steepest slide since October, on top of reporting a 67% collapse in quarterly net income. Revenue grew just 1.7%, missing estimates. Morgan Stanley cut its Tencent price target 11%.

The issue isn't the spending — Alibaba alone has committed $53B to AI over several years, and declared a $100B cloud-and-AI revenue target for five years out. The issue is that executives couldn't explain how. Bloomberg Intelligence's Catherine Lim put it cleanly: investors are waiting for "measurable revenue uplift through cloud, advertising, or transaction conversion. Until then, markets will likely stay cautious."

Why this matters beyond China: the same question — spend now, monetize later — is being asked of every AI company with a large infrastructure bill and no clear revenue ramp. US hyperscalers are spending $650B this year. The Alibaba/Tencent selloff is a preview of what happens when public markets run out of patience. Private markets, flush with the mega-fund capital raised this week, haven't had that reckoning yet. Read More…

Musk Misled Twitter Investors — Jury Finds

A California jury found on March 20 that Elon Musk misled Twitter investors in 2022 by tweeting about bots while attempting to exit his $44 billion acquisition agreement, contributing to an 8% drop in the stock. Potential damages could reach $2.6 billion. The implications for venture: Musk's legal exposure is a growing overhang on the companies connected to him. xAI raised $20B in January. If his personal legal drag accelerates, that's a headache for the cap tables around Grok and X infrastructure. Read More…

The White House Rewrites the AI Rulebook

The Trump administration released a four-page legislative framework on March 20 that would preempt all state AI laws in favor of a single federal standard, cap liability for AI developers whose products are misused by third parties, and create regulatory sandboxes for experimentation. David Sacks — White House AI czar and venture capitalist — drove the framework. House leadership immediately signaled support. The industry trade groups praised it nearly in unison, which tells you something about who wrote it. The full analysis is in the Feature below. Read More…

On March 20, the White House released a four-page legislative framework for a national AI policy. It's the most significant federal AI document since the Biden administration's 2023 executive order — and it reads like it was written by the industry it's supposed to govern.

The framework asks Congress to do six things: preempt state AI laws that impose "undue burdens," eliminate open-ended liability for AI developers whose products are misused by third parties, create regulatory sandboxes for experimentation, streamline data center permitting, avoid creating any new federal AI oversight body, and establish a single national compliance standard in place of the current patchwork. The child safety language is notably soft — responsibility shifts toward parents rather than platforms, and the framework doesn't expand existing protections like COPPA. The AI industry trade groups praised it almost in unison. Patrick Hedger of NetChoice called it evidence the White House knows "what it will take to win the future." That kind of framing tells you how the sausage was made.

David Sacks — White House AI czar, former PayPal executive, and venture capitalist — was the document's primary architect. His fingerprints are visible throughout: the anti-liability provisions mirror arguments he made publicly for years as an investor; the "minimally burdensome" framing is accelerationist doctrine dressed in policy language. House leadership immediately offered support, which matters because the Senate is less certain.

For founders and VCs, the most valuable provision isn't the one getting headlines. The preemption language matters — uniform compliance is genuinely valuable — but the liability cap is the structural win. Right now, if an AI company's product causes harm through how a third party uses it, the legal exposure is murky in most jurisdictions. Courts haven't settled it, states are moving in different directions, and plaintiff lawyers are circling. A federal framework that explicitly limits "open-ended liability" would freeze that litigation landscape in the founders' favor. That's real money.

The compliance fragmentation point is also legitimate and underappreciated. A company deploying an AI-powered hiring tool currently faces distinct disclosure requirements in Illinois, different bias-testing mandates in New York, and separate algorithmic transparency rules in California. Multiply that across 50 states with more legislation in the pipeline, and you have a legal overhead problem that disproportionately hurts smaller startups without compliance teams. A national floor removes that arbitrage.

The skeptical read is simple: this document is not a law. Congress has killed federal AI preemption twice in the last year alone — stripped from the GOP budget reconciliation bill last summer and never incorporated into the defense authorization act. The Senate is less aligned than the House leadership. The Attorney General's AI Litigation Task Force, created by executive order in December to challenge state AI laws in court, was supposed to publish its list of "onerous" state regulations by March 11. As of this writing, that list hasn't appeared. And states are not waiting — several attorneys general are already preparing resistance.

The geopolitical framing is worth noting separately. The framework explicitly argues that "a patchwork of conflicting state laws would undermine American innovation and our ability to lead" in the AI race. That framing positions opposition to federal preemption as a gift to China — a politically difficult place for any lawmaker to stand. The EU's decision to delay its AI Act high-risk obligations from 2026 to 2027 gives the administration a useful data point: even Europe blinked on aggressive AI regulation.

Position-dependent takeaway: if you're backing AI infrastructure or frontier model companies, a lighter federal touch is unambiguously favorable. If you're backing consumer AI or health AI that relies on state-level data protections as a competitive moat — or if your product genuinely depends on algorithmic bias guardrails to maintain enterprise trust — the framework cuts both ways. A national standard may weaken incumbent positioning without adding the clarity it promises. Watch which state AG challenges the AI Litigation Task Force first. That's where the real regulatory map gets drawn.

Cloaked — $375M Series B General Catalyst and Liberty City Ventures led a $375M Series B and growth financing for Cloaked, a consumer privacy platform that generates virtual identities, removes data from brokers, and screens AI-powered scam calls. The company has 350,000+ users and grew 10x over the past year. General Catalyst's Customer Value Fund provided non-dilutive growth financing alongside the equity raise — a structure designed to accelerate customer acquisition without further dilution. Security spending as AI risk hedge: the round reflects growing enterprise appetite for tools that protect against AI-generated scams, not just AI itself.

Frore Systems — $143M Series D Frore Systems, a San Jose-based developer of solid-state cooling architecture for AI compute and networking hardware, closed a $143M Series D at a $1.64B valuation led by MVP Ventures. The company's AirJet chips use ultrasonic vibrations to cool chips without moving parts or fans — directly addressing thermal constraints that limit AI chip performance density in data center and edge deployments. Infrastructure layer, not application layer: exactly the kind of picks-and-shovels bet that's dominated funding concentration in Q1 2026.

XBOW — $120M Series C Seattle-based XBOW raised $120M in a Series C led by DFJ Growth and Northzone, valuing the autonomous offensive security startup at over $1B. The platform uses AI to continuously discover and validate software vulnerabilities at machine speed. The round included Altimeter, Alkeon Capital, Sofina, and existing investor Sequoia Capital. The thesis: AI has eliminated the talent constraint for attackers. XBOW is trying to eliminate it for defenders. Timing the round to RSAC 2026 is not an accident.

Oasis Security — $120M Series B Oasis Security, developer of "agentic access management" tools, raised $120M backed by Sequoia Capital, Accel, Craft Ventures, and Cyberstarts. The company's thesis: machine identities now outnumber human users 82-to-1 in enterprise environments, and existing identity security systems were built for humans. As AI agents proliferate and request access to sensitive infrastructure, Oasis grants permissions dynamically based on real-time context rather than standing credentials.

Latent Health — $80M Series A Spark Capital led an $80M Series A in Latent Health, an AI-powered healthcare platform targeting clinical operations. Goldman Sachs Growth Equity has now backed three consecutive health AI rounds in recent weeks — Latent, Grow Therapy, and Fieldguide — confirming regulated professional services AI as an institutional-grade investment category. Details on Latent's specific product and ARR were not disclosed.

Paraform — $40M Series B Scale Venture Partners led a $40M Series B for Paraform, a recruiting marketplace that connects companies with freelance recruiters on a per-hire fee model. The structure matters: per-hire pricing aligns incentives better than per-seat SaaS, and the company is targeting technical roles where recruiter expertise is hardest to replicate with AI. Interesting bet in a market where AI is simultaneously threatening and accelerating recruiter workflows.

Paramount Skydance / Warner Bros. Discovery — $110B Merger The most consequential media deal of the decade is moving toward shareholder vote this spring. Paramount Skydance agreed in late February to acquire Warner Bros. Discovery for $31/share cash ($81B equity, $110B enterprise value including debt). The combined entity would unite HBO Max, DC, Harry Potter, and CNN with Paramount's streaming assets, NFL on CBS, and Mission: Impossible. Regulatory clearance and WBD shareholder approval remain required, with closing expected Q3 2026.

The VC angle is the broader signal: when media giants consolidate this aggressively, content distribution relationships shift, and the startups sitting in those distribution stacks — recommendation engines, ad tech, content intelligence — are repriced accordingly.

RSA Conference kicks off in San Francisco on April 28 — but with XBOW already timing its $120M raise to the event calendar, expect a wave of cybersecurity announcements to front-run the show in the coming weeks. At least two additional security startups are reportedly finalizing rounds above $75M. If three $100M+ security rounds print in a four-week window, that's not coincidence — it's a category moment.

On the regulatory front, the Commerce Department's evaluation of "onerous" state AI laws was due in mid-March and has yet to be published. That list, when it drops, will be the first concrete signal of which state laws the administration's AI Litigation Task Force actually plans to challenge. States with pending algorithmic transparency bills — New York, Colorado, and Texas among them — are watching closely. Expect immediate legal countermoves from state AGs once the list is public.

General Catalyst's $10B fundraise, if it closes at or near target, would reset LP expectations about how much capital one firm can absorb. Spark Capital's reported $3B raise would be its own benchmark. Watch for whether LP appetite is as strong as the headline numbers suggest — or whether the targets quietly compress as fundraising conversations mature. The Information first reported the Spark discussions; Bloomberg broke the GC story. Both remain early-stage and unconfirmed.