Connecticut enacts nine-platform crackdown on social media firms

Connecticut's nine-platform crackdown exposes federal preemption fault line

William Tong sent cease-and-desist letters to nine platforms on the same day. Underdog filed a federal lawsuit in response. That sequence is the story.

The Connecticut Department of Consumer Protection's order covered Underdog, Polymarket, Coinbase, Crypto.com, Robinhood, Prophet X, Novig, Webull, and Gemini — every meaningful player in the space, served simultaneously. The stated violations: allowing users under 21, accepting wagers from people on self-exclusion lists, and permitting bets on in-state college teams. Connecticut is not arguing about jurisdiction in the abstract. It is arguing about specific populations its gaming laws were written to protect.

Underdog's 39-page complaint argues that federal preemption forecloses the question entirely. The core position: Underdog operates as a federally regulated designated contract market, the CFTC has sole authority over event contracts traded on DCMs, and Connecticut cannot apply state gaming law to instruments that federal law governs. The complaint asks for both a declaratory judgment and a permanent injunction.

Robinhood has filed a parallel preemption action in the same district. Two platforms, same theory, same court, different procedural postures. That is not accident — it is coordination, or at minimum convergence on the strongest available argument.

I have seen this structure before. When multiple parties arrive at the same federal forum with the same preemption theory against the same state agency, one of two things happens: the court consolidates and rules once, or the rulings diverge and create the circuit conflict that eventually forces Supreme Court resolution. Connecticut's decision to move against nine platforms at once accelerated that timeline significantly.

The preemption argument is textually strong. Section 2(c)(ii) of the Commodity Exchange Act grants the CFTC jurisdiction over retail commodity transactions, and DCM-listed contracts occupy a specific statutory space. But Connecticut is not conceding the classification. Its position is that these instruments are functionally wagers regardless of what the CFTC calls them — and that the populations its gaming laws protect are real people with real exposure right now, not abstractions to be resolved after years of federal litigation.

That tension is where the legal risk actually lives. Courts that have looked at this closely — including the Ninth Circuit panel that ruled against Kalshi on tribal lands — have shown willingness to look through the product label to the underlying transaction. Judge McKeown called Kalshi's sports contracts a bet. A Connecticut federal judge deciding the Underdog injunction motion will have that language available.

The Montana settlement adds another data point. Kalshi dropped its Montana lawsuit and accepted a standstill: no enforcement until the Ninth Circuit resolves the en banc petition, with 30 days' notice before any state action resumes. Montana got what it needed without a final ruling. Kalshi preserved optionality. Neither side wanted to be the test case on bad facts.

Connecticut is different. Tong moved against nine platforms at once, which means he is either confident in the legal theory or willing to let federal courts resolve it at scale. The scope of the order suggests the latter. A regulator uncertain of his ground does not send letters to nine companies on the same day.

The enforceability of Connecticut's position depends on a question the district court has not yet been asked to answer cleanly: whether a DCM designation by the CFTC preempts state consumer-protection law as applied to specific populations — minors, self-excluded gamblers — that the CEA does not address. The Ninth Circuit's tribal analysis turned on IGRA's explicit prohibition. Connecticut's consumer-protection theory has no equivalent statutory anchor. That asymmetry is where the platforms have their strongest ground, and where I think the preliminary injunction motion is more likely than not to succeed.