A new bill tries to outlaw 'jawboning' — the quiet way DC moderates your platform

What happened

The Electronic Frontier Foundation is backing a new federal bill aimed squarely at jawboning — the practice of executive branch officials informally pressuring social media platforms to remove or suppress content that is, by every available legal standard, protected speech. The bill, introduced this week, would create a private right of action allowing users whose lawful posts were taken down after government contact to sue the responsible official, and would require platforms above a size threshold to log and disclose communications from federal officials about specific user content.

The framing is narrow and deliberate. The bill does not touch court orders, lawful subpoenas, or content that is itself illegal (CSAM, true threats, sanctions evasion); it targets the murky middle where an FBI liaison, a White House staffer, or a CDC official emails a trust-and-safety lead and 'flags' a post for 'review.' Platforms remain free to moderate however they want under Section 230 — the bill regulates the government side of the conversation, not the platform side.

This is the legislative echo of *Murthy v. Missouri*, the 2024 Supreme Court case where Louisiana and Missouri sued over the Biden administration's COVID-era contacts with Facebook, Twitter, and YouTube. The Court ruled 6–3 that the plaintiffs lacked standing — not that the conduct was constitutional, just that the specific users suing couldn't prove their specific takedowns traced to specific government emails. The dissent, and EFF, read that as a roadmap: build a statute that fixes the standing problem by giving any aggrieved user a direct cause of action.

Why it matters

For a decade, the dominant free-speech anxiety online has been platform power — Zuckerberg deciding what you can say, an opaque appeals process, a Trust & Safety Council nobody elected. Jawboning inverts that frame. The threat model isn't the platform deciding to remove your post; it's the platform removing your post because a federal official asked nicely, off the record, with the implicit weight of antitrust, Section 230 reform, and FTC enforcement sitting behind the 'ask.'

What makes this hard is that the pressure rarely looks like pressure. The discovery in *Murthy* surfaced emails where White House officials called platform employees 'partners,' demanded weekly takedown reports, and in one now-famous exchange asked Facebook to remove a parody account 'ASAP.' No subpoena, no warrant, no formal legal process — just a relationship, a Slack-equivalent, and the understanding that the regulator on the other end of the email has a long memory.

The EFF's argument, and it's a strong one, is that this is the textbook end-run around the First Amendment. The government cannot directly order Meta to delete a post criticizing vaccine policy. But it can, apparently, ask. And the platform, weighing the cost of a hostile FTC investigation against the cost of one user's post, will often comply. Constitutional rights that bind the government do not bind Meta — so if the government can outsource censorship to Meta, the First Amendment becomes a procedural inconvenience rather than a constraint.

Not everyone agrees the bill is necessary, or workable. Critics point out that some government-platform contact is genuinely benign and arguably required: CISA flagging foreign influence operations, the FBI warning of imminent violence, the State Department reporting sanctioned entities. A bill that chills all such contact could leave platforms blind to real threats. The drafters' answer is the carve-out for illegal content and the disclosure regime — sunshine, not silence. Whether that line holds in practice is the open question.

There's also a partisan asymmetry worth naming. The loudest jawboning complaints of the last five years came from the right, about COVID and election content. But the mechanism is bipartisan — the Trump administration's pressure on platforms over Hunter Biden coverage, on TikTok, on January 6 retrospectives, used the same playbook. A statute that constrains the practice constrains every administration, which is probably why EFF — historically allergic to coalitions of convenience — is willing to back it.

What this means for your stack

If you work in trust and safety, platform policy, or even backend engineering at a large social product, this is the bill to read carefully. The compliance surface is not 'moderate less' — it's 'log every government contact about user content, attribute every takedown to a specific reason, and be prepared to produce both in discovery.' That's a real engineering ask. Most large platforms already have internal ticketing for law enforcement requests; few have unified logging for the informal 'hey, can you take a look at this' email that lands in a policy lead's inbox.

Concretely, expect three implementation shifts if the bill (or anything resembling it) passes:

First, structured intake for government communications. Today these arrive as email, DM, phone call, in-person meeting at a conference. Tomorrow they'll need to land in a system of record with sender identity, agency, content referenced, action requested, and platform response. Think Salesforce-for-jawboning, with audit trails. Smaller platforms below the threshold get a pass, but anything Discord-sized and up will need to build or buy this.

Second, takedown attribution. Your moderation pipeline probably tags actions with a policy reason ('hate speech,' 'spam,' 'platform manipulation'). It probably does not tag the proximate cause ('CDC email dated 2024-03-12'). The bill's private right of action lives or dies on that attribution chain. Expect legal to start asking for it whether or not the bill passes — *Murthy*-style discovery is now table stakes.

Third, a chilling effect on the informal channel itself. Once officials know their 'flag' email may be served on a plaintiff's lawyer in 18 months, they'll either move to formal process or stop asking — and that, EFF would argue, is exactly the point. The risk for platforms is that the informal channel was sometimes useful, and its loss means more lawful-but-awful content stays up longer.

Looking ahead

The bill's odds of passage in this Congress are middling at best — speech regulation is hard to whip, and both parties have administrations they're trying to protect. But the direction of travel is unmistakable. *Murthy* didn't end the jawboning fight; it deferred it to a better-pleaded case or a statute. This is the statute. Whether or not it passes intact, the disclosure-and-attribution architecture it describes is going to become the de facto compliance posture for any platform with US government exposure within two years. Build the logging now; the lawyers will thank you later.