Product1 distinct publisher2 min readPublished
Stud or Dud runs liens, bankruptcies and courthouse files against a date's name. The data is the old B2B product; the new part is the red flag telling you what to make of it.
The Product Desk · Product desk
Compiled by The Product DeskSomething wrong?How this is made
Six named record and directory types go in, four labelled sections come out [17]. That compression is the product. A tax lien reads one way to a lender and another way to someone deciding whether to meet a stranger for a drink, and the report performs that translation before the buyer sees anything, then puts an AI-written summary on top of it [4][5]. A chat assistant branded "the dating detective" extends the same service to first-date advice [c6b].
What is on sale is interpretation, not access. The company's own description of its inputs is elastic by design: "public records, publicly available information, and other permitted sources" [2]. That is the phrasing you use when the list is expected to grow.
The billing model is a subscription, and Amber Higgins recommends the site as a "sidekick" to the dating apps [9]. That is positioning rather than plumbing. Nothing in the search depends on an app being involved; a name typed into a box is the whole interface [3].
The safety case, as Wired frames it, arrives with a fraud number: the FBI counts more than $40 million lost by San Francisco daters to romance scams in 2025 [13]. Hold that against the mechanism. A search keyed to a name or phone number returns records filed under that string [3], and romance scams generally start with a name that is not the operator's. The population a courthouse-and-liens report can actually describe is the population that has US records under the name they gave you, which is the ordinary local date rather than the person running a script from somewhere else. The tool is well aimed at DUIs, evictions and bankruptcies [4][7] and structurally blind to invented identities.
Higgins says modern dating has changed dramatically and that the tools people use to establish trust have not kept pace [16]. The gap she describes is real, and the answer she has shipped closes it by making a background report a consumer good priced by the month [9]. Everything underneath was already indexed and already searchable for clients running background reports [1]. The build was a front end and a verdict layer.
That is the part worth watching in the rest of the sector. Nobody has to buy new data to follow this. They have to buy an AI summariser and a payment page, and then decide which of the records they already hold deserve a coloured flag.
Ranked by verification strength, evidence, and original report placement.
Stud or Dud launched in the US and was created by the company behind PeopleFinders.com, an online data broker that collects public records from multiple sources and makes them searchable to help clients locate individuals or run background reports.
According to its website, Stud or Dud uses "public records, publicly available information, and other permitted sources" to create a profile of a potential match.
All a dater needs to run a search is the person's name or phone number.
Stud or Dud aggregates courthouse documents, bankruptcy filings, tax liens, pre-foreclosure data, phone directories, social media posts and more into a snapshot of a potential suitor's personal life.
The aggregated data is presented in a single report divided into four categories: finances, property or assets owned, employment history, and criminal background.
Follow any of these and your For You feed starts watching them — no settings page required.
Evidence-backed comparisons of source perspectives and observed adoption signals. Read the methodology
Which Builder, Operator, and Investor concerns the observed source mix emphasized—not a truth score.
Evidence, demonstrated adoption, hype gap, incentives, and confidence are assessed independently, each on its own current evidence. How these are measured.
Vendor-described product, one adversarial hands-on test
Every mechanical fact — the input feeds, four report categories, AI overview, assistant, flag layer, subscription gate — comes from the vendor's website and a CEO interview carried by a single publisher. The one independent check is Galperin's informal name-and-number testing, reported without methodology or sample size. There is no documentation, pricing page, accuracy audit, or second newsroom in the cluster, so product existence and shape are well evidenced while performance is essentially unevidenced.
Day-one US launch, zero usage data
The only hard adoption fact is that the product shipped in the US on the publication date. The company's supporting signals — dating safety as a top historical PeopleFinders use case, and a prediction that women will be earliest adopters — are unquantified assertions rather than measurements. No subscriber counts, pricing, revenue, integrations, or third-party traffic data appear anywhere in the supplied material, so the score reflects shipped-but-unmeasured.
Safety promise outruns demonstrated capability
The product is marketed as a trust-confirmation and danger-flagging tool, yet the only hands-on test in the record found it poor at exactly that, the CEO concedes no report can guarantee safety and pushes interpretation onto the user, and WIRED notes a clean report is uninformative given how underreported sexual violence is. The gap is positive and substantial: the records pipeline is real and the AI packaging is real, but the safety claim the subscription is sold on has no supporting evidence and one contradicting test. It is not maximal because the vendor does hedge publicly rather than promising certainty.
Vendor-sourced launch story with advocacy counterweight
The commercial incentive is explicit and strong: a data broker monetizing an already-owned records pipeline through a consumer subscription, with the launch narrative supplied by its own CEO and website on launch day. That is textbook promotional timing. Incentives are partly disclosed and partly offset — WIRED names the broker provenance and platforms the EFF, whose privacy-advocacy mission carries its own directional interest, plus an academic critic. Absent price and volume disclosure, the reader cannot audit the commercial claims.
Single publisher, single source, vendor-dependent
Confidence is limited by structure rather than by internal contradiction. One article from one publisher supplies the entire cluster; the descriptive claims are reliable because they restate the vendor's own public material, but nothing about accuracy, pricing, uptake, or regulatory standing can be corroborated. The critical evidence is one researcher's undocumented test. Product existence is high-confidence; nearly everything consequential is not.
product
US courts turn government hacking into a line item, starting with 2028 wiretap data1 distinct publisher
product
The UK plant that went dark for four days was too small to have to tell anyone1 distinct publisher
security
CISA finally counts the water intrusions: 100-plus exposed systems behind cellular modems2 distinct publishers
security
A North Korean IT worker got hired by a federal agency. Vetting is a security control now.1 distinct publisher
Distinct publishers with included, body-backed reporting in this cluster.
1 article · August 26, 2026