Bench IQ vs Pre/Dicta: how they compare in 2026

The models are not layered on a product, they constitute it, and they also constitute the corpus. What Bench IQ sells is a characterisation of how a named federal judge reasons, derived by large language models and AI agents from rulings that were never issued in writing. Remove the models and nothing remains that could be sold: there is no repository, no workflow, no document store, no search index over published opinions, and no dataset either, because the dataset is itself the product of machine extraction from oral rulings and other unwritten indicators rather than a collection that existed before the models were pointed at it. The privacy policy names large language models as the mechanism for analysing rulings, and the vendor describes fine-tuning commercial models on the dataset it built. This is the second A on this axis in the pull, alongside Pre/Dicta, and on the constitutive test it is the stronger of the two: Pre/Dicta applies a model to federal dockets that exist independently of it, while here the underlying record is assembled by the model in the first place.

The model is the product. A user enters a case number and receives a forecast; there is no repository, workflow, drafting surface or document store underneath that would function if the predictive model were removed. Nothing remains but the training corpus. The vendor states the position directly, describing itself as neither a self-serve analytics tool nor a case law research vendor, and the interface is built around the prediction rather than around search. This is the clearest instance of an A on this axis in either pull: not a model layered on a platform, but a platform that exists to deliver model output.

Nothing is published on accuracy or grounding for a product that produces assertions about what a named judge thinks. The site is five pages and the home page is the product page: three FAQ blocks, a social proof claim and a demo request. Checked those, the about, press, privacy policy and responsible disclosure pages on 31 Aug 2026, and there is no accuracy figure, no test set, no evaluation, no error rate, no statement of failure modes, and no hallucination disclosure of any kind. The verification question is sharper here than on any other record in this lane, and it follows from the product's own premise. Bench IQ's value proposition is that it reaches the roughly ninety-seven per cent of rulings that produce no written opinion, which means the source material behind an insight is by definition material a reader cannot pull up in a conventional research tool and check. Nothing published states whether an insight links back to the ruling it came from, whether a user can inspect that source, or what the system does when the record is thin. Two limbs of this axis do not apply rather than being failed: a citator or good-law check is not in scope for a product that characterises judicial reasoning rather than validating authority, and it is neither credited nor penalised.

The only vendor located in either pull that publishes a measured accuracy figure rather than asserting accuracy. The claim is 85 per cent on motions to dismiss, with the test set described as more than 50,000 motions to dismiss across all 94 US federal district courts, drawn from roughly two decades of federal litigation data. Failure modes are named against the vendor's own commercial interest: the 85 per cent applies to motions to dismiss alone, and for the other nine motion types Pre/Dicta states it is modelling outcomes rather than predicting them, a distinction carried onto the home page where one motion receives a prediction and every other a forecast. One inconsistency a buyer should know about and reconcile with the vendor: an interview transcript published on Pre/Dicta's own site states 87 per cent, against 85 on every other surface, and nothing reconciles the two. Two limbs of this axis do not bite on this product rather than being failed by it. Grounding to openable primary authority does not apply because the product forecasts outcomes rather than producing cited legal assertions, and citator status does not apply for the same reason. The vendor addresses the equivalent question in its own idiom by disclosing that forecasts are generated independently of case facts and legal precedent, which tells a buyer what the number does not rest on. Separately, and not a deduction under this band, the method behind the figure is not described, so whether it was produced by holdout backtest, live tracking or in-sample validation is unknown and no third-party validation was located. Checked 31 Aug 2026.

No oversight structure is published for a system whose output feeds live litigation strategy. Searched the home page, about page, press page, terms of service, privacy policy and responsible disclosure page on 31 Aug 2026: nothing describes what the models decide unattended, no review surface is documented, no confidence or abstention behaviour is described, no statement tells a lawyer what weight an insight should carry, and no route exists to report or correct an insight that proves wrong. The vendor is publicly careful about the framing, its chief executive saying the aim is not to predict the outcome of a case but to help lawyers shape it, which locates judgement with the lawyer. That distinction is real and is recorded here, but it was made in an interview rather than in product or policy material, and a framing is not a control.

A real review surface is published and the rest of the control structure is not. The vendor commits that every output is inspectable down to the underlying docket, that every underlying comparable case is visible, inspectable and linked, and that outcomes are verifiable, which gives a lawyer a genuine route from a forecast back to the matters that produced it. The platform also allows a user to re-weight or change the factors behind a prediction, which is a form of human control over model output rather than passive consumption. What is absent, checked 31 Aug 2026: nothing states that a lawyer must review before relying, no threshold or confidence gate is described, and nothing addresses what happens when a forecast proves wrong. On a product whose output feeds settlement valuation and venue strategy, the absence of any published statement about the weight a forecast should carry is the gap.

Adoption counts and anonymous testimonials stand in for evidence. The vendor publishes two figures, both self-reported and neither naming anyone: four of the top five AmLaw firms are said to use the platform, and an earlier announcement put twelve of the hundred largest US firms as pilot customers actively using the product on live cases. Three testimonials appear on the home page attributed only by seniority and firm rank. No customer is named anywhere, no case study exists, and no figure is published for what changed at any organisation. One conflation to guard against, because the same names circulate in both roles: Cooley, Fenwick & West, Wilson Sonsini and several Kirkland & Ellis partners are investors in Bench IQ, not disclosed customers. Both bands have some purchase here, since an adoption count is closer to real deployment evidence than a logo strip, and the lower is taken because the count is unverifiable, unattributed and silent on outcome. Checked the home, about and press pages on 31 Aug 2026.

One named customer of real weight, with no figures attached to it. John B. Quinn, Executive Chairman and Founding Partner of Quinn Emanuel, is quoted on the record stating that all attorneys at the firm now have access to the platform, which is a named firm-wide deployment at a major litigation practice rather than a logo or an anonymous testimonial. The vendor also holds ALM Legal Week's Innovator of the Year award. What is missing is measurement of use: no figure for what changed at Quinn Emanuel or anywhere else, no case study with a stated method, no customer count, and no other named customer located on the surfaces read on 31 Aug 2026. The 85 per cent accuracy figure is a product measurement rather than deployment evidence and is graded on the Citation Accuracy axis instead.

The published policy is a website and application privacy notice rather than a customer data commitment, and the distinction governs this grade. Effective 20 February 2025, it covers visitors to the site and users of the application, and what it addresses it addresses well: personal data limited to name, job title, employer, work email and work phone, with the vendor noting that this information is already public through firm websites, state regulators and PACER; no sale or rental of personal information; thirteen subprocessors named outright; storage and processing in the United States only; and a named security officer, the co-founder and chief technology officer. Several limbs of this axis do not apply to this product rather than being failed by it, and are neither credited nor penalised: the platform ingests judicial records rather than client documents, so there is no client matter material to segregate, no ethical wall to inherit and no privilege or work product question about ingested content. What does bite is unanswered. A litigator researching a named judge on a named issue discloses the matter they are working on, and nothing published states whether that query history is retained, analysed, used to improve the models or disclosed, and no customer-facing confidentiality commitment of any kind was located on 31 Aug 2026.

The privacy policy carries a dedicated section on case reports that addresses the actual exposure this product creates, which is rarer than it sounds. It recognises that case numbers and reports may be sensitive, and commits that Pre/Dicta does not share, sell or disclose case numbers, case report contents or access history to third parties, that encryption and strict access controls prevent unauthorised viewing, and that case access logs are used only for security monitoring and troubleshooting rather than marketing or profiling. It also states that case numbers are not stored against a personal profile unless the user creates an account. That is a considered answer to the real risk here, which is that the matters a litigator looks up reveal the firm's docket and strategy. Three gaps: no statement on whether anything a user enters trains any model, no retention period, and privilege and work product are not addressed. Note the underlying exposure is lower than for a document platform, since the input is a case number rather than client files.

Nothing published on the advice line, and no page exists that would ordinarily carry it. The site inventory was established directly rather than assumed: navigation of login, about us, careers and press, with privacy policy, responsible disclosure, security and terms of service in the footer. There is no disclaimer page and no ethics or professional responsibility page, and none of the pages read on 31 Aug 2026 states that an insight is not legal advice, that no attorney-client relationship arises, or that a lawyer remains responsible for the argument they build on it. No bar or ethics guidance is named, including ABA Formal Opinion 512, and no jurisdiction or practice limit is stated. The exposure is not incidental: the product exists to shape how counsel argues to a specific judge in a live matter. One limit to state, because it is the rebuttal route: the terms of service could be read only in fragments recovered through the search index, so a professional advice disclaimer could sit in the unread portion.

Nothing published on the advice line, and nowhere it could sit. Searched the home page, the platform page, the privacy policy and the site footer on 31 Aug 2026. There is no disclaimer, no ethics or professional responsibility page, no bar or ethics guidance named including ABA Formal Opinion 512, and no statement that a forecast is not legal advice or should not substitute for professional judgement. There is also no terms of service to carry one: the footer link labelled Terms of Conditions resolves to the home page rather than to a document, and no terms were located anywhere. The exposure is among the highest on this index, because the published use cases are settlement valuation, venue selection, whether to move to dismiss and whether to challenge an expert, all decisions a client relies on counsel to make.

No governance position is published for a system that models the reasoning of named federal judges. Searched the home, about, press, privacy policy, responsible disclosure and terms pages on 31 Aug 2026: no responsible AI statement, no accountable owner for model behaviour, no pre-release testing regime, no fairness or uneven-output evaluation, and nothing addressing what it means to characterise how identifiable individuals on the federal bench think. The method is expressly withheld, which the vendor attributes to pending patents, so a buyer cannot examine either the governance or the technique. This is the third vendor in this category with the same gap and each takes a different form, which is what makes it a pattern rather than three coincidences: Trellis displays judges' political affiliation as a data field, Pre/Dicta uses political affiliation and net worth as model inputs, and Bench IQ reconstructs judicial reasoning by large language model from material that was never published and cannot be independently checked. That last is the furthest from a verifiable public record and carries the least published governance of the three.

No governance position is published for a model whose central design choice is the one this axis exists to examine. Pre/Dicta's forecasts are driven by judicial biography, and the vendor names the variables: educational background, appointment history, political affiliation, net worth and geographic location, with judges compared against cohorts of similar background. Its own marketing has promoted comparative dismissal rates for Republican appointees, and an enhanced biographical intelligence feature lets users isolate a judge's political affiliation to see how it bears on outcomes in cases like theirs. Searched the home page, platform page, privacy policy, news index and footer on 31 Aug 2026 and located no responsible AI statement, no accountable owner, no pre-release testing regime, no fairness or bias evaluation, and no discussion of what it means to model identifiable federal judges by politics and personal wealth. A vendor may well have defensible answers here; none is published, and on this product the question is not incidental but constitutive.

Substantive published policy covering most of the ground, and one element better than anything else in this pull. The privacy notice effective 20 February 2025 names thirteen subprocessors outright, including Amazon Web Services, Heroku, Segment, Mixpanel, FullStory, Intercom, HubSpot, Notion, Slack, Sentry and Datadog. No other vendor across pull 2 publishes a list of that specificity without an access request. Alongside it: storage and processing stated as United States only; retention stated as the duration of the relationship plus a further period for the vendor's own operations and archiving; a deletion route by email; data processing agreements and model clauses with vendors described as entered into where feasible and appropriate; essential cookies only in the application; a named security officer, the co-founder and chief technology officer; and a published responsible disclosure programme with an in-scope list, a reporting address, a discretionary reward, a commitment not to pursue good-faith researchers, and a requirement to avoid privacy violations and disruption during testing. What holds this short of the top band: the post-relationship retention period is not quantified, no access control model is described, and no breach notification practice was located.

A general privacy policy covers the ground in principle without specifying any of it. It states industry-standard technical and organisational measures including encryption, secure servers and limited employee access, commits that case report access logs are automatically purged after a limited retention period unless legal obligations require longer, and addresses cross-border transfers through Standard Contractual Clauses. Searched the home page, platform page, privacy policy and footer on 31 Aug 2026 and located no security page of any kind: no encryption standard, no access control model, no named subprocessor, no incident or breach notification practice, and no stated retention period. The policy's own credibility is undercut by a defect recorded separately, which is that the document was published with unreplaced template placeholders in place of the company name and the privacy contact address.

What is published disclaims the exposure and allocates nothing toward the buyer. The terms of service provide the services on an as is, with all faults and as available basis, and disclaim all warranties express or implied including merchantability, fitness for a particular purpose and non-infringement. The surviving provisions are named and the direction is one way: the customer's obligation to indemnify Bench IQ survives termination, as do limitations on Bench IQ's liability. Bench IQ may terminate without advance notice at any time and for any reason. No indemnity running to the customer was located, no warranty attaches to any insight, and no insurance position appears. Two limits belong in the record rather than in the grade. No liability cap figure was recovered, so whether one exists is unknown. And the recovered text reads as an evaluation agreement rather than a subscription agreement, framing permitted use as testing and evaluating the services and providing feedback, which raises the question of whether a separate commercial agreement governs paying customers. Both follow from the terms being readable only in fragments through the search index on 31 Aug 2026, and both are the rebuttal route if a fuller reading changes the picture.

Nothing is published, because there is no agreement to publish it in. The footer of every page carries a link labelled Terms of Conditions, and it resolves to the home page rather than to any document. Searched for terms of service, terms and conditions and any subscription agreement on 31 Aug 2026 across the site and the wider web and located none. The consequence is total: no indemnity, no liability cap, no warranty or disclaimer of warranty on output, no insurance position, no allocation of risk of any kind between vendor and customer. This is a product that forecasts whether a motion will be granted and is marketed for settlement valuation and reserve-setting by insurers, sold without published terms addressing what happens when a forecast is wrong. Recorded as a vendor-side absence rather than a retrieval failure: the link exists and points to the wrong place.

No integration into any practice system exists, and this is an absence rather than a gap in the reading. The site inventory was established directly from the navigation and footer: login, about us, careers, press, privacy policy, responsible disclosure, security and terms of service. There is no integrations page, no developer or API documentation, and no product page on which an integration might be described, because the home page is the product page. Nothing names a document management system, matter management, e-billing, court filing, Word or Outlook, and no export or data feed is described. The product is reached through a browser at its own application and stands alone. Checked 31 Aug 2026.

No integration into any practice system was located. Searched the home page, platform page, support page link, privacy policy and footer on 31 Aug 2026. No API is offered or referenced, no document or matter management system is named, no billing or filing system connection appears, and there is no integrations page. What the vendor publishes in this area are workflow conveniences rather than integrations: the platform is cloud-based and reachable from any location, reports can be shared under the firm's own letterhead, matter numbers can be assigned to reports for cost tracking, and no technical training is required to implement. Useful to a buyer, but none of it moves data between Pre/Dicta and the systems a litigation team already runs.

The residency answer is published plainly, which is more than either incumbent in this category manages. The privacy notice states that personal information is stored and processed in the United States only, so a buyer knows where the data sits and where it is handled, and the two are not left to be inferred from one another. The subprocessor list makes the hosting identifiable rather than implied, naming Amazon Web Services and Heroku. What is not published: no tenancy model, no statement of whether customer environments are separated, no region options and no enterprise or private deployment tier, and nothing describing where inference runs as distinct from where records are stored. That is what holds it below the top band rather than any vagueness about location. Worth recording for calibration: a five-page site from a company three years old states its data residency clearly, while Docket Alarm and Bloomberg Law both sit a band lower on this axis, one of them offering two different answers across its own documents.

Cloud delivery is implied and neither the tenancy model nor the region is stated. The only infrastructure statements located on 31 Aug 2026 are that the platform is cloud-based and accessible from any location, and that data may be transferred to and processed in other countries where the vendor's servers or partners operate, with Standard Contractual Clauses and other lawful safeguards applied to those transfers. That acknowledges cross-border processing while naming no country, no region, no hosting provider and no customer-selectable option. Nothing describes single or multi-tenancy, and nothing distinguishes where data is stored from where it is processed. No residency commitment of any kind is offered.

A real, named, third-party trust centre exists and is linked from every page of the site, hosted on Drata at a public trust URL. That structure is genuine and is why this is not the floor: the vendor has stood up an independent compliance portal and points to it prominently, which is the connector that parent-level evidence elsewhere in this category lacks. What the site itself states about certification is nothing. No standard is named, no auditor identified, no scope described and no coverage period given on any page read on 31 Aug 2026. The portal's contents could not be read: it refused automated access to two independent fetchers on that date, which is recorded as a retrieval limit on the index's side under the standing rule that a page which will not extract is not an empty page, and nothing about what it contains is inferred here. This grade therefore rests only on what the vendor's own pages state, which is that a trust centre exists. The band fits imperfectly and the note should say so, since the wording at this level describes badges displayed on a site and there are none here. A proposed band amendment is logged, the second on this axis.

A badge with nothing behind it. The footer of every page carries an AICPA SOC for Service Organizations mark, in the non-CPA form licensed to service organisations, which asserts a SOC examination without stating which one. No accompanying text names SOC 1 or SOC 2, no Type I or Type II is specified, no auditor is identified, no coverage period or report date is given, no scope is described, and no route to request a report exists. There is no trust centre or security page anywhere on the property. Checked the home page, platform page, privacy policy and footer on 31 Aug 2026. A mark of this kind is meaningful only alongside the report it refers to, and the report is neither published nor offered.

The vendor identifies the class of model and never says whose. Large language models are named in the privacy notice as the mechanism by which rulings are analysed, and the company has described fine-tuning commercial large language models on its own dataset, so a buyer knows that third-party foundation models sit underneath. No provider is named, no model or version is identified, no processing location for inference is given as distinct from the stated United States storage, and no commitment to notify customers when the model or its training data changes was located on 31 Aug 2026. What makes the omission notable rather than routine is the company it keeps: the same privacy notice names thirteen subprocessors outright, covering hosting, analytics, session recording, support, marketing and observability, and not one of them is a model provider. A vendor that lists Sentry and Datadog by name and omits the systems that generate its core output has drawn a line somewhere, and nothing published explains where or why.

An unusual case: the vendor describes its own model in more detail than most of this pull and says nothing about anything it might depend on. What is disclosed is substantial for a first-party system, covering machine learning models trained on roughly two decades of federal litigation data at around 15 to 20 million cases, 50 to 100 data points per case, dozens of variables for parties and firms, and biographical profiling across hundreds of federal judges. Nothing published identifies a third-party model provider, and the architecture as described appears to be proprietary machine learning rather than a licensed foundation model, though no statement confirms that either way. Three gaps checked 31 Aug 2026: no model version or release identifier, so a buyer cannot tell which model produced a forecast or when it changed; no commitment to notify customers of changes to the model or its training data; and no subprocessor or infrastructure provider named.

No pricing information is published at any level, including the unit of charge, and no pricing page exists to carry it. The site inventory was established directly rather than assumed: the navigation is login, about us, careers and press, and the footer adds privacy policy, responsible disclosure, security and terms of service. There is no pricing page, no plan or tier names, no feature split, no per-seat, per-judge or per-report unit, no term length and no statement of what implementation adds. The single call to action on the home page is a demo request. The only adjacent statement located is a 2024 interview in which the chief executive referred to flexible pricing models and said the product is designed for firms of all sizes, which names neither a figure nor a unit and does not appear on any vendor page. Checked 31 Aug 2026.

No pricing information is published at any level, including the unit of charge. Searched the home page, platform page, privacy policy, the full primary navigation covering platform, about, research centre, news, support and contact, and the footer on 31 Aug 2026. There is no pricing page, no tier structure, no per-seat, per-case or per-report unit, no volume banding and no statement of what implementation adds. Every call to action is a demo booking or a case consultation request. The one adjacent statement is that the platform requires no technical training to implement, which speaks to effort rather than cost. Nothing published would let a buyer form any view of price before entering a sales process.

The claim on the product surface is total and the actual boundary appears only elsewhere. The home page describes a service covering all of a judge's rulings rather than the roughly three per cent producing written opinions, which reads as complete coverage, and it identifies the audience as attorneys and the buyer as a large litigation firm. The real limits are stated in funding announcements rather than on any product page: coverage is the United States federal judiciary only, the platform opened in commercial bankruptcy, and state courts, where the vendor notes the number of judges is orders of magnitude larger, are an expansion target along with other common law jurisdictions. A buyer reading only the site would not learn that their state court judges are absent. No practice area list appears anywhere, no court-by-court coverage statement exists, and nothing identifies which judges are covered or how far back. Checked the home, about and press pages and the vendor's own funding releases on 31 Aug 2026.

The buyer segmentation is more precisely reasoned than most and the coverage boundary is only half stated. Four distinct buyers are named with the question each brings: litigators setting strategy and timing motions, general counsel choosing outside counsel on fit and proven results, private equity valuing law firm acquisitions on portfolio-level performance, and insurers setting reserves against how comparable matters resolve. Each is given its own surface. Coverage is stated concretely on the federal side, spanning all 94 district courts, the appellate layer, and ten named motion types from motions to dismiss through Daubert challenges, temporary restraining orders and motions to remand. The limit is where it thins: state coverage is described only as select state courts, with no list of which, no county or court detail and no depth statement, and the January 2023 acquisition of Gavelytics brought state trial court assets whose current coverage is not published anywhere. Checked 31 Aug 2026.

Searched the home page, about page, press page, privacy notice, responsible disclosure page and the terms of service fragments recoverable through the search index on 31 Aug 2026. No located term or policy states whether anything a customer puts into the product is used to train or refine the models, either way. The exposure differs from a document platform, since the corpus is judicial records rather than client files, but it is not absent: what a litigator asks about, meaning which judge and which issue, is customer-specific information, and the privacy notice lists improving and enhancing the services among its purposes without saying whether that extends to model training on user activity. The notice governs website and application users rather than customer data under a subscription, so the question a general counsel would ask is not addressed by the document that exists.

Searched the home page, the platform page, the privacy policy in full and the site footer on 31 Aug 2026. No located term or policy addresses whether anything a customer enters is used to train or refine the forecasting model, either way. The closest statement is that the vendor improves the service based on aggregated, anonymised usage data, which covers behavioural telemetry rather than model training on customer input. Two things reduce the exposure without answering the question: the input is a case number identifying a public federal docket rather than client documents, and the privacy policy commits that case numbers, report contents and access history are not shared, sold or disclosed to third parties. No terms of service exists in which a training term could sit.

Retention is acknowledged and not quantified. The privacy notice states that information is kept only as long as necessary, and elsewhere frames the period as the duration of the relationship plus a further period for the vendor's own operations and archiving, with a deletion route offered by email. No number attaches to either limb. Nothing separately addresses queries put to the platform or the insights returned: no retention period for them, no customer-configurable setting and no zero-retention option. Read 31 Aug 2026 alongside the terms of service fragments, which carry no retention provision in the portion recoverable.

Retention is acknowledged and left unquantified. The privacy policy states that case report access logs are automatically purged after a limited retention period unless legal obligations require them to be held longer, and that personal data is retained only as long as necessary for the purposes described. No period is given for either, no retention position is stated for the case reports themselves as distinct from the access logs, and no customer-configurable setting or zero-retention option is offered. Read 31 Aug 2026.

Nothing addresses separation, and much of what this signal asks does not apply to this product. Bench IQ ingests judicial records rather than client documents, so there is no firm corpus to index, no document management access model to inherit and no matter-level material that a retrieval layer could read across. Recorded as not addressed because that is the honest value, with the reason stated: this is a question the product class does not raise in its usual form rather than one the vendor has neglected. What would apply and is unanswered is narrower: nothing published states whether one user's research activity is visible to colleagues inside a subscribing firm, which matters where two teams in the same firm are on opposite sides of a matter before the same judge. Searched the privacy notice, terms fragments, home, about and responsible disclosure pages on 31 Aug 2026.

Protection against unauthorised viewing is asserted without a described model. The privacy policy commits to encryption and strict access controls to prevent unauthorised viewing of case reports, and states that case numbers are not stored against a personal profile unless the user chooses to create an account, which is a real minimisation step. What is not published is any permission structure: nothing describes separation between customer organisations, nothing addresses access between users inside a subscribing firm, and nothing states whether saved or linked case reports are visible firm-wide. On a product where the set of case numbers a firm looks up maps directly onto its live docket, internal visibility is the question a buyer would ask, and it is unanswered.

An unusual and creditable disclosure that does not answer the question this signal asks. The privacy notice states that since it was founded the company has received zero government requests for information, which is a transparency report in miniature and the only published count of its kind located across pull 2. What it does not do is commit to anything: nothing states what would happen if a request arrived, whether the customer would be told, whether the vendor would seek a waiver where notice is prohibited, or whether it would resist. Recorded as not addressed on that basis, since the value set turns on notice rather than on history, and the disclosure is carried here so the record does not read as though nothing were published. Checked the privacy notice, the terms fragments and the responsible disclosure page on 31 Aug 2026.

Disclosure to authorities is permitted and notice is not addressed. The privacy policy states that limited data may be shared with legal authorities if required by law, court order or government request. No commitment to notify the customer accompanies it, no undertaking to seek a waiver where notice is prohibited, no minimisation obligation and no transparency report. That sits against a stronger commitment elsewhere in the same document, that case numbers, report contents and access history are not disclosed to third parties, and nothing reconciles the two. Whether a firm would learn that its access history had been demanded therefore rests entirely with the vendor. Read 31 Aug 2026.

Coverage is described by jurisdiction and the corpus behind it is deliberately undisclosed. What is published: the United States federal judiciary, opening in commercial bankruptcy, with the material said to span the roughly ninety-seven per cent of rulings that produce no written opinion, including rulings delivered orally from the bench and other indicators. What is not published is how any of that was obtained. The vendor states that details of the dataset and the technique behind it are confidential because patents are pending, so no source is named, no licence or public-record basis is given, no supplier is identified and no update cadence is stated. The privacy notice mentions PACER, but as a public source of attorney contact details rather than as a corpus source. This is the sharpest instance of the provenance question in the pull, because the corpus is both the entire differentiator and the one thing the vendor has decided not to describe. Checked 31 Aug 2026.

The corpus is described by type and volume in more detail than most of this pull, and its rights basis is not stated anywhere. Published figures cover roughly two decades of federal litigation data at around 20 million classified cases and 40 million judicial decisions, with 50 to 100 data points per case and several million parties and firms, drawn from federal dockets across all 94 district courts. A second corpus goes undiscussed as such: the judicial biographical data driving the forecasts, covering educational background, appointment history, political affiliation, net worth and geographic location for hundreds of federal judges. No source is named for that material, no licence or public-record basis is given for either corpus, and no update cadence or data vintage date was located on 31 Aug 2026.

Searched the home, about, press, privacy notice and terms surfaces on 31 Aug 2026. No citator, treatment signal or currency check is claimed, and none applies: the product characterises how a judge has reasoned rather than retrieving authority whose standing needs validating, so subsequent history is not a question it raises. Recorded as not addressed because that is the honest value, with the reason stated rather than left to look like neglect. One adjacent currency question does bite and is unanswered: nothing published states how far back the judicial record runs, how often it is refreshed, or what happens to a judge's profile when they move court or leave the bench.

Searched the home page, the platform page, the privacy policy and the footer on 31 Aug 2026. Nothing addresses whether authority surfaced through the platform is checked for subsequent history, and no citator or treatment signal is claimed. The product forecasts outcomes rather than retrieving law, and the vendor states that its forecasts are generated independently of legal precedent, so the question sits outside what it sells. One feature is adjacent: Precedent Intelligence surfaces prior matters selected by compositional similarity rather than citation, and the vendor notes these often fall outside the scope of traditional research tools. No currency or good-law check is described for them.

Searched the home page, which is also the product page, along with the about, press, privacy notice, responsible disclosure and terms surfaces on 31 Aug 2026. Nothing documents what the platform does when the record on a judge is thin, when a judge is newly appointed, or when an issue has not come before them. No abstention path is described, no confidence or coverage indicator is mentioned, and no statement addresses how a user would know the difference between a well-evidenced characterisation and a sparse one. The question has unusual weight on this product: its premise is that it reaches material nobody else has, so a user has no independent way to gauge how much sat behind a given insight.

Searched the home page, the platform page, the privacy policy and the footer on 31 Aug 2026. No explicit no-answer or abstention path is documented, and nothing states what the platform does when comparable cases are too few to support a forecast, which matters given that state court coverage is described only as select. Two features sit adjacent to this without meeting it. The vendor draws a published distinction between a prediction, offered only for motions to dismiss, and a forecast for the other nine motion types, which tells a user where confidence is lower. And output for the modelled motions is expressed as outcome likelihoods, grant rates and denial patterns rather than as a binary answer. Both are calibration disclosures rather than an abstention behaviour.

No court order, opinion or disciplinary record naming this product has been located as of 31 Aug 2026. Instrument searched: the AI Hallucination Cases database maintained by Damien Charlotin, which tracks decisions worldwide where a court addressed hallucinated AI content and records the tool implicated where known, searched on the product name and the unspaced variant, alongside 2026 sanctions trackers and law firm commentary indexes. This is a statement about the public record on the date shown rather than a clearance, and it is bounded by what that database covers. The failure mode this signal tracks fits the product only loosely: Bench IQ characterises judicial reasoning rather than generating citations to authority, so a fabricated citation reaching a filing would have to originate elsewhere, though an inaccurate characterisation of a judge is a distinct risk this signal does not capture.

No court order, opinion or disciplinary record naming this product has been located as of 31 Aug 2026. Instrument searched: the AI Hallucination Cases database maintained by Damien Charlotin, which tracks decisions worldwide where a court addressed hallucinated AI content and records the tool implicated where known, searched on both the product name and the unspaced variant, alongside 2026 sanctions trackers, law firm commentary indexes and trade press summaries. This is a statement about the public record on the date shown rather than a clearance, and it is bounded by what that database covers. The product's output is a statistical forecast supported by real dockets the user can open rather than generated citations to authority, so the failure mode this signal tracks is not the one the product is exposed to.

Searched the full site inventory on 31 Aug 2026, established directly from the navigation and footer rather than assumed: home, about us, careers, press, privacy notice, responsible disclosure, security and terms of service. No bar or ethics authority is engaged with anywhere, including ABA Formal Opinion 512 and any state bar guidance, and no page maps the product to a professional conduct obligation. There is no disclaimer page and no ethics page in the inventory at all, so the absence is structural rather than an omission from a page that exists. Nothing addresses the questions a risk committee would raise about a tool that profiles the judge before whom the firm appears.

Searched the home page, the platform page, the privacy policy, the full primary navigation and the footer on 31 Aug 2026. No engagement with any bar or ethics guidance was located, including ABA Formal Opinion 512 and any state bar material. Nothing addresses the professional conduct dimension of the product at all, which is notable on a tool built to inform whether to move to dismiss, how to value settlement and which venue to choose, and one that models judicial behaviour by political affiliation. No terms of service exists in which such a statement could otherwise sit.

Searched the home, about, press, privacy notice and terms surfaces on 31 Aug 2026. No per matter record of AI-assisted work exists, no guidance on billing, fee or client disclosure treatment is published, and no time-saving claim was located either, which is why this records as not addressed rather than as savings claims only. The vendor's public framing runs to outcome rather than efficiency: its chief executive describes litigators crafting smarter strategies and delivering better results clients will pay for. That is a claim about the value of the work rather than about the hours behind it, and it engages neither limb of this signal.

Public materials are framed around cost control and predictability rather than time saved: more accurate litigation budgets, alternative fee arrangements and cost forecasting, clearer visibility into litigation risk and exposure, more competitive responses to requests for proposal, and optimised resource allocation. One feature is closer to this signal than most vendors get: reports can be assigned matter numbers to simplify cost tracking, which supports allocating the tool's cost to a client matter. Searched the same surfaces on 31 Aug 2026 and located no per matter record of AI-assisted work intended for fee purposes and no guidance on billing, fee or client disclosure treatment where a forecast informs advice.

A current subprocessor list is published openly, with no agreement or access request required, and it is the most specific in pull 2. The privacy notice names thirteen subprocessors outright, among them Amazon Web Services, Heroku, Segment, Mixpanel, FullStory, Intercom, HubSpot, Notion, Slack, Sentry and Datadog, and states that data processing agreements and model clauses are entered into with vendors where feasible and appropriate. A firm asked by a client which third parties sit behind the tool has a real answer to forward, which almost nothing else in this category offers without a portal request. Two gaps keep it from the top value. No model provider appears on the list, so the question of which system generates the insights is unanswered by the very document that answers everything else. And no client-facing disclosure or consent material is published to accompany it.

Searched the home page, the platform page, the privacy policy and the footer on 31 Aug 2026. Nothing that would support a client-side disclosure obligation was located: no subprocessor list, no model provider or infrastructure provider named, no trust centre, no data processing agreement, no named certification behind the SOC mark in the footer, and no client-facing consent or notification material. The privacy policy refers to service providers covering cloud hosting, analytics and payment processing under confidentiality terms, which acknowledges third parties in the chain without identifying any of them. A firm asked by a client which vendors see its matter information would find nothing to answer with.

Searched the home page, about page, press page, privacy notice, responsible disclosure page and terms surfaces on 31 Aug 2026. Nothing addresses court disclosure or verification certification. No model is identified or versioned, so which system produced a given characterisation cannot be established; no record of sources behind an insight is described, and the underlying rulings are by the product's own premise mostly unwritten, so there is no opinion a lawyer could attach; nothing records who reviewed an insight; and no export of any kind is offered. A lawyer under a standing order requiring disclosure of AI use, or certification that outputs were checked, would find nothing here to build the answer from.

The sources behind an output are traceable and the AI provenance is not. The vendor commits that every output is inspectable down to the underlying docket, that each comparable case is visible, inspectable and linked to its docket, and that outcomes are verifiable, which means a lawyer can produce the real matters a forecast rests on. Reports can also be shared under the firm's own letterhead, which is a distribution feature rather than a disclosure record and arguably works against provenance by stripping the vendor's mark. Two elements are missing: no model is identified or versioned anywhere, so which system produced a given forecast cannot be established, and no export designed for a court disclosure or AI-use certification was located on 31 Aug 2026.