Think of Us as Google Reviews for Public Officials.

We’ll soon be launching a YouTube channel that does one thing: runs elected officials’ public social media accounts through REVOLT Insights LeX™ and shows you what the algorithm finds.

No legal opinions. No allegations. No interaction with any official’s account.
Like camping, we’ll only leave gentle footprints in the log file.

We find posts.
On X, Faceboo, Insta, Blue Sky.
We score them.
We show you the ones that scored high.
The rest is yours to interpret.

Also, this is all public information, so we’ll be publishing names and feeds.

The channel launches with two episodes built around a single question: what does the difference between a clean account and a flagged one actually look like in data?

The answer is two heat maps. They tell the story before we say a word.

The Problem

Two years after the unanimous Supreme Court decision in Lindke v. Freed (601 U.S. 187, 2024), most elected officials have never had a single social media post reviewed against the two-prong state-action test.

The standard municipal guidance — don’t delete, don’t block — is the entire institutional response to a ruling that exposed every official with a mixed-use account to potential 42 U.S.C. §1983 liability.

NLC, IMLA, and NACo published excellent work after the ruling — webinars, FAQs, educational materials. Their own guidance acknowledges cities should consult general counsel when implementing policy. Most haven’t. The water cooler guidance is still the state of the art.

The chain of command — PIO, City Manager, City Attorney — has a gap exactly where the liability lives. Nobody is watching the feeds. Nobody pre-approves moderation decisions. When a thin-skinned official gets a comment they don’t like at 11pm on a Tuesday, nobody is in that loop.

What REVOLT Insights LeX™ Does

REVOLT Insights LeX is a classifier built by REVOLT Insights that scores public social media posts by elected officials on a 0–100 scale across Authority, Duty, and Perception factors — the same framework the Court used in Lindke to distinguish personal speech from government speech.

Eventually, we’ll publish under a Creative Common License and put into the public domain, for review and improvement.

The review threshold is 60. Below it: personal content, minimal state-action indicators. Above it: posts that raise the question.

What we are not doing:

• Not interacting with any official’s account
• Not telling officials what to do
• Not saying “this is illegal”

We surface patterns. Courts (Federal and Public Opinions) and counsel will decide.

How LeX Is Built: The STAN Stack

LeX doesn’t run on a single AI model. It runs on five.

REVOLT Insights uses a multi-model adversarial methodology called STAN — where one model writes the briefs, some build, some destroy, and a locally-trained model that knows our voice does the final pass. A red team attacks the output. A green team builds it. A validator flags what can’t be verified.

The result is output that has been argued over before it ever surfaces. Hallucinations get caught in the pipeline. What remains survived scrutiny.

Four explicit human checkpoints. Every score verified. Every post 3x checked. AI drafts. Humans decide.

The results train the model … it’s self-improving.

Everyone else uses AI to finish sentences. We use it to start fights.

The Green Team: Deerfield Beach Commissioner Tom Plaut

Deerfield Beach District 4 Commissioner Plaut’s Facebook feed scored an average of around 20/100 across several hundred posts.

That is not close to the line.
That is what a personal account looks like.
Plaut does it right.

Post distribution by Lindke Exposure Tier — Commissioner Plaut. Mass concentration at 0–19. Nothing above the review threshold. Consistent across the entire post sequence.

The heat map confirms it. No behavioral shifts, no spikes, no anomalies. There is a structural tell that reinforces the score: Plaut explicitly directs followers to a separate official page for commissioner activity. He drew the line himself. LeX confirmed it holds.

Moderation risk: minimal — no state-action indicators sufficient to trigger Lindke analysis
– REVOLT Insights LeX

LeX says:

• Posts above review threshold (60+): none
• Execution time: < 10 seconds
• Score tier: Low-Mod — mostly personal with minor governmental signals
• Heat map: mass concentration in 0–19 bands, flat and consistent across entire post sequence, no behavioral shifts or anomalies
• Structural finding: account explicitly directs followers to a separate official commissioner page — self-imposed boundary between personal and governmental communication
• Moderation risk: minimal — no state-action indicators sufficient to trigger Lindke analysis
• Conclusion: clean personal account pattern with incidental civic references
• Account does not appear to announce government decisions, respond to constituent service requests, direct residents to city departments, or moderate policy debate about municipal matters

For the record — I think he is a good guy. He is just not up to speed on the specifics, and neither is almost anyone else in local government. I have submitted public records requests to municipalities across the board and have yet to find a single city that has folded Lindke compliance into written, enforced social media guidelines.

Plaut is the green team. His numbers are fine.

Disclaimer: I’ve never exchanged a word with Plaut, nor was he aware of our socials scan or this post.

The Red Team: A Different Official, a Known Bad Actor

Now look at this.

Post distribution by Lindke Exposure Tier — anonymous official. Sustained density in the 60–89 bands across 19,000+ posts. Not a spike. A pattern.

Same tool. Same threshold line. Different data set.
Simply put, darker boxes above 60+ have raised LeX’s algorithmic interest.

Sustained medium-to-dark density in the 60–69, 70–79, and 80–89 bands running across the entire post sequence. Not a spike. Not an anomaly. A pattern. This account posts above the threshold the way Plaut posts below it — consistently, persistently, across thousands of posts.

Score tier: High to Critical — consistent concentration in 60–89 bands across all post windows
– REVOLT Insights LeX

LeX said:

• Account analyzed: anonymous elected official, platform withheld
• Posts reviewed: 3,500+
• Execution time: < 2 minutes
• Posts above review threshold (60+): sustained density throughout entire post sequence
• Score tier: High to Critical — consistent concentration in 60–89 bands across all post windows
• Heat map: medium-to-dark density above threshold line running the full length of the sequence — not a spike, not an anomaly, a pattern
• Behavioral pattern: account posts above threshold the way Plaut posts below it — consistently, persistently, at scale
• Structural finding: no visible separation between personal and official communication — mixed-use account with no self-imposed boundary
• Score bands most active: 60–69, 70–79, 80–89 — sustained across all windows
• Moderation risk: significant — sustained state-action indicators across thousands of posts create ongoing exposure under Lindke two-prong test
• Conclusion: account exhibits behavioral pattern consistent with government speech function — posts regularly exhibit Authority, Duty, and Perception indicators above the review threshold

LeX flagged these posts as high boundary-indicator density. The official’s own words are in those cells. The only person who put that content above the threshold line is the official who posted it. We’ve anonymized their information, but you know their name.

They out themselves. We built the index.

We didn’t serve you that terrible sandwich, we just warehouse the review.

The Ghost of Christmas Future: Deerfield Beach Resident Dan Herz

If Deerfield Beach resident, and perennially campaigner, Dan Herz wins Deerfield Beach’s District 1 seat, the structural conditions for a §1983 Civil Rights claim exist before he finishes his first term. This is not unique to Herz — any thin-skinned official, like Herz, dropped into the same structural vacuum behaves the same way under pressure. The conditions are the story, not the personality.

Here is the checklist that completes itself the day any such official is sworn in:

Thin-skinned officials do not invent new behavior under pressure. They do what they have always done — delete the comment, block the account, make the problem disappear. Except now the problem does not disappear. Now it has a case number.

If your FB page is how you communicate legislative matters, and you block me, how am I supposed to know what’s going on, how do I redress my grievances?  That official used a grenade, and not a scalpel.

Constituent posts something the official does not like on a high-scoring post. Official does what thin-skinned people do. Constituent screenshots it. LeX already scored the post above threshold before any of this happened. The §1983 complaint is three paragraphs. The §1988 attorney fee clock starts running.

That is a six-figure legal bill over the horizon. Not a worst-case scenario — the actuarial outcome of the current setup, applied to a specific personality type, in a specific seat, with no institutional guardrails.

The city handed them the liability and called it onboarding.

Who Owns This?

Nobody. That is the answer.

The PIO thinks it is the City Attorney’s problem. The City Attorney thinks it is the PIO’s problem. The City Manager thinks someone built a policy. Nobody built a policy. The chain of command has a gap exactly where the liability lives.

REVOLT Insights built LeX to do the indexing governments cannot do themselves. Our client base is municipalities. Citizens surface the targets. Cities fix the problem or become the next episode. We did not build it to file lawsuits. We built it so cities would not have to defend themselves.

Two Ways to Engage

Cities choose which one they would rather be on.

Been Blocked?

If an elected official has blocked you, deleted your comment, or hidden your reply on a social media account they use for government business — we want to know about it.

Drop us an email. Tell us who and where. We will run the account through LeX and see what the data says.

Your free tip is how we find the next account to run. If the data flags it, the municipality hears from us, but we will not share with you.

We are not a pathway to filing lawsuits.

Want to Know Your Numbers?

We are offering three ways in.

Free scan — Send us one official’s public social media handle. We run LeX and return the tier breakdown. No cost, no commitment. 48-hour turnaround. First 10 cities only. Not for personal or corporate sales.

Single official audit — $3,500 — Full LeX scan, heat map, top flagged posts surfaced, account structure assessment. Everything your City Attorney needs to start the conversation. Deliverable in 15 business days.

Second official scan included at no additional charge when you purchase a single official audit. Two heat maps, one decision. See the range inside your own municipality before anyone else does.

Cities, County Attorneys, Risk Managers, Municipal League members — this is the baseline you do not have. Get it before your next election cycle closes the window.

Contact: chazstevens@revolt.training

I am not a lawyer. Nothing here is legal advice. Qualified legal counsel is usually a good idea. LeX surfaces posts for consideration. All conclusions are the reader’s own.

FAQ

Q: What is Lindke v. Freed?

A unanimous 2024 Supreme Court decision establishing when a public official’s social media account is subject to the First Amendment. If an official posts government speech on a personal account and moderates comments, they may face 42 U.S.C. §1983 liability.

Q: What is REVOLT Insights LeX™ and how does it work?

REVOLT Insights LeX™ is a classifier built by REVOLT Insights that scores public posts by elected officials on a 0–100 scale using the same Authority/Duty/Perception framework the Court used in Lindke. Posts above 60 are flagged and surfaced. No legal opinions. Just the data.

Q: Can a city be sued over an official’s social media?

Potentially yes, under Monell doctrine, if no written policy or training exists and a pattern of violations occurs. REVOLT Insights is not a law firm. This is not legal advice. Consult your municipal attorney.

Q: What does a clean score look like?

Density concentrated at 0–39, nothing sustained above the 60-point threshold. Commissioner Plaut’s account averaged approximately 20/100 — a documented clean result.

Q: What does a flagged score look like?

Sustained density in the 60–89 bands across the post sequence. Not a spike — a pattern. The heat map shows it clearly.

Q: Public channel vs. private audit — what’s the difference?

Public channel findings are published openly. Private audits run under attorney-client privilege and NDA through the city’s outside counsel — nothing surfaces publicly.

Q: Is this legal advice?

No. REVOLT Insights is not a law firm. LeX surfaces data. Legal conclusions belong to qualified counsel.

Q: I was blocked by an elected official. What do I do?

Email chazstevens@revolt.training with the official’s name and jurisdiction. We will run LeX and see what the data says. We will only share the results with the municipality.