Darrow
Two Supreme Court clerks set out to find people who didn't know they'd been harmed.
Verdict: Darrow's founders left Israel's Supreme Court to find hidden victims, and its homepage now sells litigation risk.
A Narrative Authenticity Framework analysis of Darrow. NAF score 8.0 out of 10. Weakest component: Real Aspiration, at 6.5.
Evyatar Ben Artzi and Elad Spiegelman clerked together for Justice Uzi Vogelman at Israel's Supreme Court. What they saw there bothered them. People harmed by real violations often never reached a courtroom at all.
"Most people don't even know that they were the victims of a violation," Ben Artzi told NoCamels in 2022. The two clerks left the court and thought about starting a nonprofit.
Then they met Gila Hayat, a data scientist who had served in Unit 8200, the Israeli army's signals intelligence unit. Hayat saw a data problem. The evidence sat in public records, and nobody had a way to find it and connect it.
The three founded Darrow in 2020. Its AI reads public data such as consumer complaints, filings and news, spots likely violations, estimates how many people were harmed, and brings the case to a plaintiffs' firm.
That's a justice story, and darrow.ai has stopped telling it.
Where Darrow stands
Darrow raised $35 million in September 2023, TechCrunch reported. The homepage now lists more than $22 billion in legal exposure surfaced, more than 22,000 legal professionals using its intelligence, and logos from 30 law firms, including Milberg, Baron & Budd and Bernstein Litowitz.
In May 2026 Darrow launched a platform that lets plaintiffs' firms manage litigation like an investment portfolio. LawNext's headline called Darrow the "AI Lab for Legal Risk."
Why the brand line falls short
The homepage opens with "Legal Risk Doesn't Begin with a Lawsuit." The About page calls Darrow "an AI research lab." Neither page mentions the clerks, the Supreme Court or the nonprofit they nearly started.
The word justice survives in one heading, "Justice Through Intelligence." On the page for law firms, harmed people appear as figures on a sample case card: 113,000 victims and a $55 million attorney's fee.
The shift makes commercial sense. Darrow sells to law firms and litigation funders, and they buy returns. The founders left the court over the person who never knew they'd been wronged, and the new language leaves that person out.
The NAF scorecard
| Component | Score | One-line read |
|---|---|---|
| Friction | 8.5 | A real buyer problem, named on the homepage |
| Lived Stake | 8.5 | Strong founding story, told only in press |
| Distinctive Qualification | 8.5 | $22 billion surfaced and 30 named firms |
| Real Aspiration | 6.5 | Justice language traded for portfolio language |
| Overall | 8.0 |
That puts Darrow level with Addepar, Airwallex and Qdrant on the SeekArgus leaderboard. Patagonia (9.5) is the benchmark.
Scores as of publication. Current standings at seekargus.com/leaderboard.
Friction: 8.5
Darrow names a real problem in its own primary copy: by the time litigation starts, "the window to act has already closed." Law firms and corporate legal teams feel that. The older friction, harm that nobody ever sees, was sharper and belonged to Darrow alone.
Lived Stake: 8.5
Few founding teams sit this close to the problem: two clerks at a country's highest court and a data scientist from an army intelligence unit. NoCamels and Calcalist have told the story, and Darrow's newsroom reposts that coverage. Darrow's About page doesn't tell it, which keeps the score below 9.
Distinctive Qualification: 8.5
The proof is specific: $22 billion in exposure surfaced, 30 named law firms, and underwriting that Darrow says draws on more than 105,000 cases. What's missing is a public record of outcomes showing which cases settled and who got paid.
Real Aspiration: 6.5
Darrow used to describe itself as a "Justice Intelligence Platform." That named a changed world, where harm gets found and the people harmed get compensated. The current brand describes a better tool for managing legal exposure, which is a capability.
What a 9 would look like: the homepage names one group of people Darrow's work helped compensate, with the amount and the date, and says that this is why the company exists.
The differentiated story
Darrow's most distinctive story is the person who never knew they were a victim. Its data now shows it can find that person at scale, and its brand has stopped mentioning them.
We clerked at Israel's Supreme Court and watched harm go unseen because no one could connect the evidence. So we built a system that reads public data the way a lawyer would. Every case we surface starts with people who didn't know they'd been wronged.
Three moves to raise the score
1. Put the clerks' story on Darrow's own About page (Lived Stake 8.5 to 9.0)
Tell the Supreme Court origin in the founders' voice: what they saw, the nonprofit they nearly started, the meeting with Hayat. Link it from the homepage. Measure time on page and how often press coverage quotes the founders' account.
Prior-work check: Partial. Darrow's newsroom reposts press coverage of the founding, and the About page links to the Darrow Creed. Neither tells the story in Darrow's own voice.
2. Publish a record of settled cases and who got paid (Real Aspiration 6.5 to 8.0)
A dated public list of surfaced cases that have settled, with the harm, the number of people covered and the amount returned to them. Measure how often press and prospective firms cite it.
Prior-work check: Partial. The law-firms page shows a sample case card with expected settlement and victim counts, and Darrow says many surfaced cases have settled. No public record of actual outcomes found.
3. Explain in public how Darrow gets paid (Friction 8.5 to 9.0)
Sacra reports that Darrow can share in attorney fees through Arizona-based lawyers. That model will draw questions from courts, clients and reporters. A plain page on how Darrow earns money and handles conflicts answers them first. Measure how often fees come up as an objection in sales calls.
Prior-work check: Partial. The integrity value on Darrow's About page commits to "the highest ethical standards." No public explanation of the fee model found.
Projected composite: 8.6 (Friction 9.0, Lived Stake 9.0, Distinctive Qualification 8.5, Real Aspiration 8.0). Checks cover public sources only, as of Nov. 19, 2026.