📡 The Big Story
The FCC wants to publish a Robocall Scorecard, a public, provider-by-provider report card grading how well every carrier and intermediary actually blocks illegal traffic. Not a filing. Not a compliance attestation buried in the Robocall Mitigation Database that nobody reads. A scorecard. With your name on it. That the press can screenshot.
Here's why this matters more than the fifteenth STIR/SHAKEN rulemaking you've ignored: it changes the incentive structure completely. Right now, robocall mitigation is a cost center. You do the minimum, you file the attestation, you move on. Nobody buys your termination service because you block 4% more junk than the next guy. But the moment there's a public number next to your company name, mitigation becomes a sales asset for the good actors and a procurement red flag for everyone else. Enterprise compliance teams are going to start pasting that scorecard into vendor reviews. I guarantee it.
The part I'm genuinely watching is methodology, because that's where this gets ugly. Block rate alone is a garbage metric. A provider that blocks aggressively looks great on a scorecard and terrible to the legitimate businesses whose appointment reminders got eaten. False positives are the real story in this industry and they're much harder to measure than "calls stopped." If the FCC grades on volume blocked without weighting wrongful blocking, we get a race to over-filter, and the collateral damage lands on exactly the legitimate A2P traffic everyone claims to want to protect. Get the methodology right and this is the most useful thing the Commission has done in years. Get it wrong and we've built a leaderboard that rewards being trigger-happy.
🔥 What's Moving
FCC Cuts Off 14 Providers From U.S. Phone Networks 💀
The Commission blocked fourteen companies from connecting to U.S. phone networks over Robocall Mitigation Database violations. Not a fine. Not a consent decree. Disconnected. For a voice or messaging intermediary, that isn't a penalty, that's a death certificate.
And look at the timing next to the scorecard proposal. Carrot in one hand, guillotine in the other. The message to every aggregator and reseller is unambiguous: your RMD filing is not paperwork, it's your operating license. If you have a compliance person who "gets to it when they get to it," this is the week to change that. Fourteen companies just found out what the deadline actually was.
Palm Beach Tan Pays $2.5M for Texting After STOP 🔥
A tanning chain is now the most important TCPA case of the quarter. Palm Beach Tan settled for $2.5 million over texts sent after consumers opted out, and the settlement includes something rarer than the money: an unusually explicit list of what counts as a valid stop request.
Here's the thing that gets companies sued. Your platform honors STOP, UNSUBSCRIBE, CANCEL, END, QUIT. Great. Your customer replies "please stop texting me" or "take me off this list" or "STOP." with a period, and your regex shrugs. A human read that as an opt-out. A jury will too. Every SMS program reading this should spend an afternoon running that keyword list against your own opt-out logic, plus fuzzy variants, punctuation, and casing. $2.5 million is an expensive way to learn your parser was too literal.
GrapheneOS Is Building Its Own RCS Stack 👀
GrapheneOS announced it's adding RCS to its own Messaging app to reduce dependence on Google Messages, alongside a redesign and a secure paste feature that blocks apps from reading other apps' clipboard data.
Small user base, enormous question. RCS Universal Profile is supposedly an open GSMA standard, but in practice on Android it's been Google Messages and the Jibe backend, full stop. If a small privacy-focused team can actually ship an independent, interoperable client, the standard is real. If they hit a wall of undocumented dependencies and proprietary provisioning, then everyone who has been quietly calling RCS "a Google channel with a standards body attached" gets to say so out loud. For anyone building business messaging on RCS, that answer determines whether you're adopting a protocol or a platform.
Google Messages Sent Old Texts to the Wrong People 🤡
Users on Pixel and Samsung devices reported that old messages from unrelated conversations resurfaced and got delivered alongside new texts and images. Google confirmed a fix is in progress. The bug surfaced September 2.
Misdelivery is the worst class of messaging bug because there's no undo. And the timing is almost too perfect: this is the same app the entire industry is being told to route brand RCS traffic through, in the same week an alternative client project got announced. Nobody asked for a live demo of why client diversity matters, but here we are.
🏆 Winner of the Week: Compliance teams, who spent years being the department that says no and just got a $2.5M settlement and fourteen dead providers as exhibits A and B.
📉 Loser of the Week: Google Messages, which shipped a privacy bug and an existential competitor in the same seven days.
📊 By the Numbers
- 14 providers blocked from U.S. phone networks in a single action. That's not a warning shot, that's the FCC demonstrating it will actually pull the plug. Every one of those companies had a filing obligation they treated as optional.
- $2.5 million for sending texts after STOP. Divide that by the cost of an engineer spending one week hardening your opt-out parser and the ROI is genuinely absurd.
- 1 independent RCS client attempt, from a team with no carrier deals and no billion-user install base. The fact that GrapheneOS is the first serious try, in 2026, tells you everything about how open this "open standard" has been.
🔮 What We're Watching
The scorecard comment period. This is where the fight happens, and it will be about methodology, not principle. Nobody will say "we oppose consumer protection." They'll say "we have concerns about metric design," which is lobbyist for "please don't measure the thing we're bad at." Watch which carriers push hardest for aggregate stats over per-provider detail. That tells you who's nervous.
Whether GrapheneOS actually ships working RCS. Not the announcement, the ship date. If an independent client sends and receives Universal Profile messages against carrier networks without Google's blessing, the RCS conversation changes permanently. If it stalls, that's an answer too.
💡 The Hot Take
Public scorecards are going to do more to clean up American phone networks than a decade of STIR/SHAKEN did, and it's not close.
Technical mandates get gamed. Attestation levels became a checkbox roughly eleven minutes after they existed. But reputation is not gameable, because the audience isn't the regulator, it's your customers' procurement departments. Here's my actual prediction: within eighteen months of the first scorecard publication, at least one major enterprise brand publicly cites provider rankings as the reason it switched aggregators, and the messaging industry starts quietly putting mitigation performance in sales decks the way cloud vendors put uptime SLAs in theirs.
And the bolder half of the take: the same treatment is coming for A2P messaging. Voice got the scorecard first because the robocall problem is older and louder, but text fraud volume is heading the same direction, and regulators have now seen that publishing numbers works. If you run a messaging platform, assume your filtering performance becomes public within three years and start building like it's already true. The companies that treat compliance as a product feature instead of a legal cost are about to look very smart, and the ones optimizing for whatever the scorecard happens to measure are going to block a lot of appointment reminders finding out.
