Bloomberg Law vs Blue J: how they compare in 2026

The machine learning is deep, long-standing and genuinely load-bearing for several tools, and the product it serves is a legal publishing business that stands without it. The vendor's own framing settles it: its approach page lists differentiated content as one of four pillars and says plainly that the AI does not generate answers on its own but looks through a large collection of vetted information. Strip the models out and 15.5 million court opinions, 200 million dockets, 5 million statutes and regulations, 8,200 practical guidance documents and a legal newsroom remain, which is the product Bloomberg has sold since 2009. What would not survive is a real set of tools rather than a veneer, and that is why this is not lower: BCITE, Docket Key, Points of Law and Litigation Analytics are built by machine learning and do not exist without it. The flat all-inclusive subscription reinforces the reading, since a buyer does not purchase the AI separately or meter it. Graded level with Docket Alarm and one band below UniCourt, where the machine-derived dataset is the whole deliverable rather than an access layer over one.

The models are the engine of the product a buyer uses, and a real content asset sits beneath them. Blue J states plainly that it does not create or train its own generative models, using large language technology from OpenAI and Google, so the differentiator it owns is the curated tax database rather than the model. That database is sold as a feature in its own right: the individual plan leads with full US federal and state tax coverage, the feature matrix lists case law, primary and secondary source materials and an authoritative database updated daily as line items, and the content includes licensed Tax Notes and IBFD material. Strip out the AI and what remains is a licensed tax content collection, which is a product this market already buys from Checkpoint and CCH. That is the B band: models as the engine of the core capability, layered on something that would still function as a document system. Verified 2 September 2026.

Three of the four things this axis asks for are published and the fourth, the measurement, is not. Grounding is documented with an architecture attached: the vendor names retrieval-augmented generation alongside a proprietary guardrail service, states that generated content is grounded in its own expert-written or primary source material rather than the internet at large, and says Bloomberg Law Answers provides citations and links to the supporting authorities used in producing an answer, with inline citations offering one-click verification. Citation status is checked, which almost nothing else in this pull can say: BCITE is the platform's own citator and reports whether a holding is still valid law. The system states when it has no support, abstaining where no content covers the topic or the query is outside legal scope. What is absent is any number. The page describes an extensive benchmarking process with former attorneys evaluating accuracy from development through beta and into the life of the product, and no result, test set, sample size or error rate from that process was located on 31 Aug 2026. The gap is sharper than a simple omission, because the same page commits the vendor to providing customers with clear documentation of the benchmarking and validation process for its AI, and no such documentation was found. To the vendor's credit it states directly that AI responses are never fully trustworthy and pose a hallucination risk, and advises customers to validate answers against primary sources.

Grounding is real, documented and unusually well built out at the interface. Every answer carries inline citations and a source list, and the feature matrix adds affordances most competitors do not describe: the relevant section of each source is highlighted, a user can ask for more sources, and a user can put questions to a single source in isolation. The corpus behind it is named rather than gestured at, combining primary tax authority with licensed Tax Notes and IBFD content, and it is stated to be updated daily. What is absent is measurement. No accuracy figure, test set, benchmark or evaluation result appears on any Blue J surface. One negative finding belongs on the record because a buyer will meet the number: a 90 per cent accuracy claim circulates in third-party software directories and review sites, attached to Blue J's earlier outcome-prediction product rather than to the current research platform. It was searched for on 2 September 2026 and could not be confirmed on any Blue J page, so it is recorded here as unverified and is not treated as evidence. The vendor's own subprocessor table does disclose ValsAI for LLM response benchmark testing, so testing exists; no results from it are published. Verified 2 September 2026.

The design posture is stated in a sentence most vendors would not write: the vendor says it wants its AI to make the job easier, not to do it for the user, and describes a deliberate building-block approach addressing one task at a time rather than producing the end work product. Real control surfaces sit behind that. Answers carry inline citations with one-click verification to the source, the abstention behaviour is documented with two named triggers, all generative tools are beta tested before release, and customers are advised to validate responses against primary sources and use their own judgement. What is not published, checked across the approach-to-AI page, the AI product page and both agreements on 31 Aug 2026: nothing describes what happens after an answer proves wrong, no correction or error-reporting route is documented, and no confidence threshold or configuration is exposed. The gap matters most on Deep Thinking, described by the vendor as automatically planning and executing multi-step research across its content before synthesising results, which is the least supervised thing on the platform and has no described checkpoint between the request and the finished answer.

The review obligation is contractual and the verification surface is strong. Section 7 of the Subscriber Agreement makes it the express responsibility of the subscriber and its users to review and determine the suitability of any output and to consult their own independent legal, tax and accounting advisors before using it, and section 12 of the Terms of Use frames the platform as an information tool only. The privacy notice adds a specific negative under GDPR Article 22, stating that automated decision-making of the kind that provision governs does not take place. On the product side the checking affordances are real: inline citations, source lists, highlighted source passages and the ability to interrogate a single source. What is not published is any account of what the system does on its own. No modes, thresholds or escalation conditions are described, and nothing states what happens after an answer is wrong. Autonomy is inherently modest for a research tool, but the boundary is asserted by disclaimer rather than described. Verified 2 September 2026.

Testimonials stand in for measurement. The pricing page carries three quotes, and one is better than most in this pull because it names an organisation and describes an outcome: a corporate counsel at Synovos, Inc. says the platform let the team pull state summaries of laws and regulations without calling outside counsel, reducing tens of calls on minor issues and so reducing annual legal fees. That is a real deployment account with a real effect, and it has no figure, no timeframe and no date attached. The other two are attributed only by role and segment, a practice area leader in small law and a partner in large law, and both speak to the flat-fee pricing rather than to outcomes. Searched the home page, the AI product page, the approach-to-AI page, the pricing page and the site navigation on 31 Aug 2026 and located no case study with a stated method, no customer count, no named law firm, and no quantified result of any kind. The solutions pages for law firms, in-house, government and law schools were not opened and are the most likely home for anything stronger.

The named base is deep. Roughly twenty firms are identified across the home page and customer material, including Crowe, RSM US, Adams Brown, GreerWalker, Saville, Ketel Thorstenson, Larson Gross, Barnes Dennig, Bartlett Pringle and Wolf, ELO CPAs, HMV CPAs, AGT CPAs, MBE CPAs, Perelson Weiner, Sorren, MMB and Co and the National Association of Tax Professionals, most with a named individual and job title speaking on the record. Figures are published too: three hours saved per user per week, 75 per cent less time spent on research, more than 70 per cent of users logging in weekly, and one named partner describing four to five hours of work completed in fifteen minutes. What holds it below A is that the figures and the names sit apart. The percentages are aggregate claims across the user base rather than measured outcomes at any named firm, no dates appear, and the only stated basis is that the savings calculator reflects average savings observed among users. Verified 2 September 2026.

The documents are detailed, readable before signing, and what they disclose about research activity is the concern rather than the reassurance. The privacy policy states that the information collected includes user content, usage information, browsing history and search history, along with reading habits and inferences drawn to create a profile reflecting a user's preferences and behaviour. It then describes an advertising regime over that material: creation of lookalike or matched audiences for targeted advertising on social platforms, sharing of a hashed email address with third-party advertising companies, and Sharing for cross-contextual behavioural advertising under California and Virginia law, with opt-outs offered. The policy applies to the marketing site and to the products together and does not say whether in-product search history is carved out of any of it, which is the ambiguity a litigator would need resolved, because the cases and statutes a lawyer searches disclose the matter. Alongside that, the subscription terms state that the site is not designed or intended as a repository of personally identifiable or health-related information and give no warranty as to how such information would be held, and reserve a right to audit and monitor use of the services. Real commitments do exist and are graded here: a subpoena notice undertaking, Standard Contractual Clauses and an intra-group data protection agreement, data centres stated to be dedicated to the vendor's own operations. What is missing is the whole confidentiality core: no statement either way on training with customer input, no retention period, no segregation model between users or matters, and no privilege or work product treatment anywhere.

The third-party model provider limb is answered better here than by almost any record in this pull. Blue J states that it is opted out of abuse monitoring with both OpenAI and Google and that it holds supplementary signed agreements with both prohibiting them from training models on any data coming from Blue J, and that no customer files are used for training. Retention and deletion are addressed concretely, with uploaded files older than 24 hours deleted automatically and a right to be forgotten under which any client, present or past, can have all their data permanently erased on written request with confirmation. A mutual confidentiality article sits at section 10 of the Terms of Use. Two limbs fail. Nothing published describes segregation between customers, users or matters: no tenancy or isolation statement was located on any surface. And nothing addresses privilege or work product treatment, the only mention being that no solicitor-client relationship arises with Blue J, which is a different question. The training limb is also qualified rather than clean, for the reason set out on the training signal. Verified 2 September 2026.

A clear advice-line position, published twice and reinforced outside the contract. Both the website terms and the subscription terms of service carry a capitalised clause stating that none of the services constitutes actual legal advice, opinion or recommendation, that a competent professional should be sought where legal assistance is required, that the user assumes all responsibility for decisions, conclusions and opinions reached as a result of using the services, and that no attorney-client relationship is formed with the vendor or any content supplier. Who may use the product is defined rather than assumed: the terms require a user to be eighteen, a member of a professional, business or academic community whose primary purpose is professional research, working in a court, government agency, corporation or similar entity, or enrolled as a law or business student, and exclude bots outright. The competence and supervision dimension appears where a buyer will actually read it, on the public approach-to-AI page, which advises customers to validate responses against primary sources and use their best judgement and says the AI is meant to make the job easier rather than do it. Two things hold this short of the top: no bar or ethics guidance is named anywhere, including ABA Formal Opinion 512, and no jurisdictional limit on the product's answers is stated.

The advice line is stated twice, in both operative documents, and stated well. Section 7 of the Subscriber Agreement records that Blue J is not engaged in rendering legal, tax, accounting or other professional advice, that the product is neither intended nor authorised as a substitute for the knowledge, expertise, skill and judgment of a lawyer, accountant or other professional advisor, and that users must review the suitability of output and consult their own advisors before using it. Section 12 of the Terms of Use repeats it in capitals and adds that no solicitor-client relationship is formed. Naming accountants alongside lawyers is honest given who actually buys this product. The competence and supervision dimension is therefore reached. Two limbs fail: no jurisdiction limits are stated for advice purposes, the three-country split being a commercial and coverage division rather than a professional one, and no bar, law society or professional body guidance is engaged anywhere, with neither ABA Formal Opinion 512 nor AICPA standards named. Verified 2 September 2026.

A published framework with a described mechanism behind it, and no results and no bias work at all. The approach-to-AI page, last modified 18 June 2026, sets out four pillars and puts substance under the accuracy one: retrieval-augmented generation plus a proprietary guardrail service, generative output grounded in vetted first-party or primary material rather than the open internet, beta testing of every generative tool before release, and an accuracy evaluation carried out with former attorneys running from development through internal testing, beta and customer testing and onward through the life of the product, with continuous tuning against model drift. It commits to documenting where and when the products use generative AI, the sources the AI uses, and the benchmarking and validation process. That is a mechanism a buyer can ask about, not a set of principles. What is not there is the other half of this axis. No accountable owner inside the vendor is named. No testing result is published, so the benchmarking commitment is unredeemed by anything located on 31 Aug 2026. And nothing addresses bias or uneven output, across matter types, jurisdictions or populations, on a platform whose analytics profile named judges, attorneys and firms. The trust centre carries a document titled AI Frequently Asked Questions whose contents were not read and which is the most likely place for any of this.

Governance exists as security governance and is documented well in that register, without ever reaching model behaviour. A named executive owns it, with security compliance stated to be overseen by CTO Brett Janssen, which is more accountability than most records in this pull offer. Fifteen information security policies are listed by name and reviewed annually, whole-corporation risk assessments run annually using the Cloud Security Alliance CAIQ, and Drata continuously monitors more than a hundred controls. The subprocessor table discloses ValsAI for LLM response benchmark testing, which shows a model evaluation regime exists. What is missing is everything about that regime and everything about fairness. No governance framework for AI is published, no owner of model behaviour as distinct from security is named, nothing describes what is tested before a change ships or what the benchmark measures, no certification such as ISO 42001 is claimed, and nothing at all is published about uneven output across taxpayer types, entity structures or populations. Verified 2 September 2026.

Most of the ground is covered in public, with the retention period the notable hole. The privacy policy, updated 9 July 2026, states that personal information is kept on secure servers under administrative, technical, personnel and physical measures, and sets out how long it is held by naming the criteria rather than a number: the time needed for the collection purpose, whether the user has stopped using the products, the existence of relevant legal proceedings, and legal or regulatory requirements. Its intra-group statement is more specific than most, describing a personal data governance programme, a maintained data map inventory documenting flows within the vendor's systems and disclosures to third-party systems, ongoing monitoring and testing of safeguards, company-wide privacy training, and data centres stated to be dedicated solely to the vendor's own products and operations with monitored access controls. Standard Contractual Clauses and an intra-group data protection agreement cover cross-border transfer, and additional processing terms including SCCs are made available where the vendor acts as processor. The trust centre adds data erasure, backups, access monitoring, least privilege and logging as named control areas, and publishes a full responsible disclosure policy with a reporting address and stated expectations on both sides. Missing: no retention period for anything, no subprocessor named on any page read, and no breach notification commitment located.

All five elements are published, current and specific enough to hold the vendor to, and the subprocessor disclosure is the most complete located in this pull. Thirty-one named entities appear in a table with description and location, covering not just infrastructure but the AI layer specifically: OpenAI for large language model AI, Google for LLM and AI services, Microsoft for cloud AI services, Pinecone for vector database services, Elastic Cloud for search, Exa Labs for web data search and ValsAI for benchmark testing, plus both Blue J affiliates identified as subprocessors in their own right. Retention is stated with a period where it matters, uploaded files older than 24 hours being deleted automatically, and removal of customer data is available on request through a named route. Deletion is unusually strong, with a right to be forgotten open to any client present or past and a confirmation email on completion. Access control covers enforced two-factor authentication, least privilege, annual access reviews, Auth0 identity management and BastionZero remote access with SSH disabled. Incident practice is a maintained and annually tested response plan with a commitment to report promptly to required parties. Encryption is AES-256 at rest under AWS KMS with TLS 1.2 minimum in transit, backups run daily with five-minute incrementals, and third-party penetration testing is annual with quarterly vulnerability scanning. The one gap is that no fixed breach notification deadline to the customer is published. Verified 2 September 2026.

The allocation is published in full, is precise, and runs almost entirely one way. The subscription terms of service, updated November 2025, cap aggregate liability at 500 dollars regardless of the form of action, disclaim all express and implied warranties including merchantability, fitness for a particular purpose and non-infringement, provide the services as is and as available, and state that the vendor does not guarantee the accuracy, correctness, completeness or timeliness of the services. One exclusion is written for this product class specifically and is worth a buyer's attention: liability is excluded for any inability or failure to perform legal or other research or related work, or to perform it properly or completely, even where assisted by the vendor. The terms also place the burden squarely, saying the user is solely responsible for the accuracy and adequacy of the data used and the resulting output, and impose a one-year limitation period on any action. Governing law is Virginia with exclusive jurisdiction in Arlington County. The only carve-out from the cap is death or personal injury from negligence where the law forbids limiting it. Nothing runs toward the customer: no indemnity, no warranty on output, no AI-specific liability term. The trust centre lists a certificate of liability document, which was not read and is the one route to an insurance position.

The allocation is published in full, is specific, and leaves the buyer with essentially nothing. Section 13 of the Terms of Use disclaims all warranties expressly including accuracy, reliability, currency and completeness, states that Blue J will not be liable for damages of any kind including direct damages, and then caps total aggregate liability at ten US dollars or the local currency equivalent, adding that multiple claims do not increase the ceiling. Against an individual plan priced at 1,498 dollars a year, that cap is under one per cent of a single user's annual fee, and it is the lowest located anywhere in this pull. Section 14 runs an indemnity from the customer to Blue J, expressly extending to the customer's use of the platform in connection with any legal activities, and no indemnity runs the other way. One genuine protection is published and is worth naming: the Tax Analysts terms reproduced in section 13 commit that Tax Analysts will defend against good faith, reasonable claims that use of the licensed content infringes another party's rights, so intellectual property risk on the licensed corpus is covered by the content licensor. B rather than A because nothing here is a recourse a buyer can invoke when an answer is wrong; B rather than C because a complete warranty, limitation, indemnity and content-licensor structure is published rather than a bare limitation clause. Verified 2 September 2026.

Integrations are named in passing and nowhere documented. Two are identifiable: Bloomberg Law Contract Solutions is described in the vendor's own newsroom as integrating with Microsoft Word alongside a centralised contract repository, and the subscription terms carry a special provision for users who reach Bloomberg Law through the Bloomberg Professional service, meaning the Terminal, whose own agreement then governs. Beyond that, nothing. Checked the full primary navigation covering products, solutions, insights, events and about, the AI product page, the approach-to-AI page and the site footer on 31 Aug 2026: there is no integrations page, no developer or API documentation, and no named connection to a document management system, matter management, e-billing or court filing. Dashboard Legal is offered as a project management and collaboration product rather than as a bridge into systems a firm already runs. This is the least well evidenced row on this record and the limit is worth naming: the Contract Solutions and workflow tools product pages were not opened, and either could carry integration detail that would move this grade.

No integration into the systems tax and legal work already lives in was located, and none is claimed. Checked the home page, how it works, the pricing page and its full feature matrix, the security page, the Subscriber Agreement, the Terms of Use and the privacy notice on 2 September 2026. There is no integrations page, no named connector, no API or developer documentation, and no mention of document management, practice management, tax preparation or workflow software of any kind. What exists is adjacent but is not integration: answers can be downloaded as PDF and DOCX, which is export, and single sign-on is offered with custom SSO options on the team plan, which is authentication. This matters more than it might for a research tool, because tax work runs through preparation and workflow software and a 2023 company announcement described planned API integrations that would securely leverage firm data; nothing on any current surface indicates that shipped. Verified 2 September 2026.

One infrastructure statement, no product deployment position. The privacy policy's intra-group data protection statement says the vendor's data centres are dedicated solely to its own products, services and operations and have secure and monitored access controls, which is more than the cloud-by-implication most of this lane offers and is the reason this is not lower. Cross-border movement is acknowledged rather than hidden: the policy states that uses and disclosures may involve transferring and processing personal information in various countries with differing levels of protection, including the United States, and that Standard Contractual Clauses and an intra-group agreement identifying exporters and importers are used as safeguards. What is not published anywhere located on 31 Aug 2026: no region or residency option for a customer to select, no tenancy model for the platform, no distinction between where processing happens and where data is stored, and no statement of where the generative features run. The trust centre lists a data centre item whose contents were not read. The statements that do exist sit in an annex about personal data protection rather than in any product or security material a buyer would be pointed to.

Residency is answered completely and unusually honestly, including the negatives. The security page states that all data is stored and processed in the United States in the AWS us-east-1 North Virginia region, and then states plainly that Blue J does not currently offer the ability to store data in any other jurisdiction and does not currently offer an on-premise solution. Publishing what is not available, rather than leaving a buyer to infer it from silence, is the best-evidenced form of a restriction. Hosting is named as Cloudflare and AWS, the whole subprocessor estate is listed as United States, and the privacy notice discloses that personal data of EEA, Swiss and UK individuals may be processed in the United States under transfer safeguards. What keeps this off the top band is tenancy. Nothing published describes whether the platform is multi-tenant, single-tenant or how one customer's data is isolated from another's, which is the limb this axis pairs with region and which the two European records in this pull both cleared. Verified 2 September 2026.

A trust centre that is easy to find and thin on the one detail that makes an attestation an attestation. Every Bloomberg Law page footer links to it directly, under the label Trust and Security, with a product identifier in the URL, which is a materially better position than the unlinked parent portals seen elsewhere in this lane. The portal is a SafeBase instance reachable without a sales conversation and its document set is the richest located in this pull: a penetration test report, a security whitepaper, SOC 1 and SOC 2, a SIG Lite self-assessment, a certificate of liability, a business continuity management system document and an AI frequently asked questions document, alongside named control areas across product security, access control, application security, network security, business continuity and training, and a responsible disclosure policy published in full. What is missing is the substance of the attestation. SOC 1 and SOC 2 are named without a Type, no auditor is identified, no coverage period or report date is given, and no scope statement was located. Documents are split into public and private with an access request whose tier the portal does not state. One retrieval note: the overview section returned on fetch on 31 Aug 2026 described the group's tax and accounting software rather than Bloomberg Law, most likely because the product selector resolves in the browser, so it is recorded as a limit rather than graded.

Certification is real, current and described with more process detail than most: an independent auditor maintains a SOC 2 Type 2 report on an annual basis, Drata's platform continuously monitors more than a hundred controls with automated evidence collection, third-party penetration testing runs annually and third-party vulnerability scanning quarterly. A trust centre exists and is linked by full URL rather than merely named. The access tier is established on the page itself rather than needing to be inferred, and it is the reason this is not an A: Blue J states that it can provide its SOC 2 report to customers upon receipt of a signed non-disclosure agreement, so the evidence sits behind a negotiated document rather than being reachable by a prospect. Nor is the surrounding detail published: no auditor is named, no report date or coverage period is given, and the trust services criteria actually covered are not stated. The linked trust centre was not opened this pass, but because the NDA gate is stated in terms, opening it would not change the tier. Verified 2 September 2026.

The architecture is described and the providers are not, which is the shape this band exists for. What is published: retrieval-augmented generation named as the framework, a proprietary guardrail service named as the layer over it, an explicit statement that generation is grounded in the vendor's own expert-written and primary source content rather than the internet at large, and a statement that its large language models train on that specialised content. The vendor also describes the wider technology base, naming machine learning, natural language processing, information retrieval, recommendation systems and large language models, and says Bloomberg Industry Group employs hundreds of AI researchers and engineers who publish research annually. What a buyer cannot learn from anything located on 31 Aug 2026: whether any third-party foundation model sits underneath and whose, since the phrase used throughout is our large language models without stating they are first-party; no model name or version; no processing location for inference; no subprocessor named; and no commitment to notify customers when the model or its training data changes. The trust centre carries a subprocessors item and an AI FAQ, neither of which was read, and either could name what is missing here.

The providers are named and the architecture around them is described in real detail. Blue J states that it does not create or train its own generative models and uses large language technology from OpenAI and Google, and the subprocessor table adds Microsoft for cloud AI services, Pinecone for vector database services, Elastic Cloud for search infrastructure and Exa Labs for web data search, which together describe a retrieval architecture rather than just a model dependency. Where it runs is given: AWS us-east-1, with every listed subprocessor located in the United States. The commitments about those providers are specific, with abuse monitoring opted out at both OpenAI and Google and signed agreements prohibiting either from training on Blue J data. Two of the four things the top band asks for are absent. No model is named anywhere on Blue J's own surfaces, only the houses they come from, so a buyer cannot tell which model answers their question or when it changes. And no commitment to notify customers of supply chain or model changes was located. Verified 2 September 2026.

The shape of the deal is published and the number is not. The charging model is stated clearly and repeatedly, and it is real information in a market where competitors meter: one platform at one price, flat and all-inclusive, with unlimited access rather than per-search or per-document charges. The pricing page makes the consequence explicit through a customer quote, that flat all-inclusive pricing means occasional use of a tool carries no additional expense, and the request form splits the purchase into two named lines, Bloomberg Law Research and Bloomberg Law Dashboard Legal. That is where disclosure stops. Searched the pricing page, the getting-started entry, the AI product page and the full navigation on 31 Aug 2026 and located no figure at any level, no unit of charge, no tier names or feature splits, no term length and no statement of what implementation adds. Every route to a number is a form: the primary navigation item is Request Pricing and the page it opens is a lead capture form. The firm-size bands in that form, from solo through fifty or more practitioners, imply price varies with headcount, but they are form fields rather than a published band structure.

This is the strongest commercial disclosure in the pull and it sits at the top of this band rather than in the next one. The individual plan is published at 1,498 US dollars per user per year with what it includes spelled out, a seven-day free trial requiring no credit card, and a purchase path that completes online without speaking to anyone. A feature matrix of roughly twenty-five rows sets out exactly what separates sole practitioner from team access, covering answers, sources, support and security lines including which tiers get a dedicated customer success manager, usage analytics and custom single sign-on. The page states there are no add-ons, no extra fees and no surprises, and the FAQ explains that plans are annual, changeable at any time, with access retained to the end of a cancelled period. What places it in this band rather than the top one is that the team tier, which is what most of the six thousand firms cited actually buy, is contact for pricing, with enterprise plans handled separately by sales. Regional pricing pages exist for Canada and the United Kingdom and were not opened; the figure recorded is the United States one. Verified 2 September 2026.

Segments are documented and the boundary is not. Four buyer segments each have a dedicated solution page, covering law firms, in-house counsel, government and law schools, and the government page states unlimited access to state and federal coverage. The organisation types the vendor sells to are enumerated in its own forms as law firm, tax and accounting firm, corporation, nonprofit or association, consulting firm, education institution and government, with firm size bands from solo to fifty or more practitioners. Content coverage is quantified with unusual precision for this lane: more than 15.5 million court opinions growing by around 50,000 a month, roughly 200 million dockets with more than 21 million associated pleadings, more than 5 million codified statutes and regulations, 75 million EDGAR filings and more than 8,200 practical guidance documents, with Docket Key classifying over 200 motion and brief types across all federal district courts and Chart Builders supporting state-by-state rule comparison. What is absent is any statement of where it stops. No practice area list with supported and unsupported areas, no jurisdiction-by-jurisdiction coverage table, no statement of state court depth, and no acknowledgement of a gap anywhere. Checked the home page, the four solution page descriptions in the navigation, the AI product page and the approach-to-AI page on 31 Aug 2026.

Segment coverage is documented precisely and the practice area is unambiguous. Four firm sizes each have a dedicated page, running from sole practitioner through local and regional to national, and four use cases are separately described as advisory, compliance, tax writing and training. Jurisdiction is handled properly rather than assumed, with separate United States, Canada and United Kingdom sites carrying their own pricing, security and contract sets, and the substantive coverage of the US product is stated as full federal and state and local tax. The practice area is tax and only tax, which the whole site makes plain. Two things are left open. Nothing addresses in-house tax departments or government use, which the four firm-type pages exclude by omission rather than by statement. And nothing states the boundary within tax: no list of what the corpus does not cover, and no statement of where the product stops. Worth recording for a buyer of this index: the named customer base is overwhelmingly accounting and tax practices rather than law firms, so the professional audience is broader than the lane implies. Verified 2 September 2026.

No located term or policy states whether customer input is used to train or refine any model, either way. The privacy policy lists improving, upgrading or enhancing the products and developing new ones among its purposes for collected information, which covers the ground without naming training. The approach-to-AI page says the vendor's large language models train on its own specialised content, meaning its court opinions, guidance and news, and says nothing about customer material. The asymmetry is explicit in the agreement: the subscription terms prohibit a customer from using product content in prompts or for training, tuning or incorporating into any language model without the vendor's written consent, and define language model broadly enough to catch any generative system. The prohibition runs one way. Read 31 Aug 2026.

Section 3 of the Terms of Use takes the licence quoted above over any data, information, records and files a user enters, expressly including all results from processing them and all compilations and derivative works, for four stated purposes of which the fourth is to improve the Blue J Platform. Section 4 of the Subscriber Agreement adds a separate permission to review all inputs and corresponding outputs for enhancing functionality and user experience. Neither names training or machine learning, but both permit unqualified use of customer content to improve the product, without any de-identification or aggregation limit, which places this in the agreement rather than on a policy page. Two things run the other way and a buyer needs them. The security page states that customer data is not used to train generative AI models, that Blue J neither creates nor trains its own models, and that signed agreements with OpenAI and Google prohibit those providers from training on any Blue J data, with abuse monitoring opted out at both. And section 3 ends with a deletion right on request, which sits oddly against a perpetual and irrevocable licence. Trial users get a narrower and revocable version of the review permission; paying subscribers do not.

Retention is addressed with criteria instead of a period. The privacy policy states that personal information is kept for as long as necessary for the described purposes and legitimate business reasons, and names four factors used to set the window: the time needed to achieve the collection purpose, whether the user has stopped using the products, the existence of relevant legal proceedings, and legal or regulatory requirements. Search history and user content are expressly within the categories collected, so questions put to the AI features fall inside this framework, but no period attaches to them and no customer-configurable or zero-retention setting is offered. The subscription terms point the other way on uploads, stating the vendor has no obligation to retain or store user generated content. Checked the privacy policy, both agreements and the approach-to-AI page on 31 Aug 2026.

One pathway carries a hard period and the main one does not. Uploaded customer files, where the optional file upload capability is enabled, are deleted automatically once older than 24 hours, which is the tightest published file retention rule located in this pull. Prompts and answers are a different matter: they persist as Threads, which is a described product feature rather than an incident of storage, and no retention period for them is published anywhere. Removal is available rather than scheduled, through a request to a dedicated customer success manager or the security address, and the security page adds a right to be forgotten under which any client present or past can have all their data permanently erased with confirmation on completion. The privacy notice states only that personal data is retained as long as necessary and may survive termination for legitimate business purposes. Retention is therefore acknowledged, one narrow period is fixed, and the window that matters for prompts and outputs is left open.

Nothing addresses walls or matter-level separation, and one published provision runs against the idea. The privacy policy states that personal information may be disclosed to the party who authorised the user's access, naming the employer or educational institution, which means a subscribing firm is a potential recipient of information about what its own users did rather than a party walled off from it. The subscription terms treat each registration as single-user and require credentials to be kept confidential, which is account hygiene rather than a permission model. Searched the privacy policy, both agreements, the trust centre landing page, the AI product page and the approach-to-AI page on 31 Aug 2026 and located no statement on separation between customer organisations, no access model within a subscribing firm, and no matter-level concept anywhere in the product.

Checked the home page, how it works, the pricing page and feature matrix, the security page in full, the Subscriber Agreement, the Terms of Use and the privacy notice on 2 September 2026. Nothing describes segregation between customers, between users inside a customer, or between matters. No tenancy or isolation statement appears anywhere, which is a notable omission on a security page that is otherwise the most detailed in this pull and covers encryption, access control, backups and subprocessors at length. What exists is adjacent and internal: least privilege and two-factor authentication govern Blue J's own staff access, and Workspaces are mentioned in passing as the unit across which email notifications support user collaboration, without any description of what a Workspace separates. Team plans offer custom access controls as a paid feature, which is stated as a line in the pricing matrix and nowhere explained.

A real notice commitment, published and qualified. The privacy policy addresses disclosure to unaffiliated third parties in response to a subpoena, court order or other legal process, and states that in such an event the vendor will use reasonable efforts to give the user reasonably prompt written notice so the user has an opportunity to comment or object and to preserve the confidential nature of the information, and to cooperate at the user's expense with the user's efforts to do so. That is the substance this signal asks for. Three qualifications a buyer should carry: the undertaking is reasonable efforts rather than an unconditional commitment, cooperation is at the user's cost, and the subscription terms contain a separate provision reserving the right to use and disclose information obtained from or input by the user as part of any legal process or as required by law, with no notice attached, which nothing reconciles. No transparency report was located on 31 Aug 2026.

The privacy notice addresses compelled disclosure directly and then commits to notice before acting. It records that Blue J and its Canadian, US and other service providers and affiliates may disclose personal information in response to a search warrant or other legally valid inquiry or order, including lawful access by governmental authorities, courts or law enforcement, and then states that where disclosure of a customer's information is required by law or court order it will promptly notify the customer prior to complying, unless prohibited by law, and will co-operate with the customer on the response. Notice before rather than after, plus an undertaking to co-operate on the response, is at the stronger end of this value. It is not the top value because no transparency report was located: nothing published records how many such requests have been received or how they were answered.

The corpus is identified by type and quantified in detail, with the rights basis unstated. Published figures cover more than 15.5 million court opinions growing by around 50,000 a month, roughly 200 million dockets with more than 21 million associated pleadings, more than 5 million codified statutes and regulations, 75 million EDGAR filings and more than 8,200 practitioner-written guidance documents, so both the primary law and the vendor's own editorial layer are described and a cadence is given for the opinions. No upstream supplier is named. The privacy policy refers only to third party data suppliers as a category, and discloses something a buyer should note in passing: the databases available to registered users contain public records, publicly available information and non-public information from those suppliers. The subscription terms acknowledge third-party content providers whose own rules attach to their material without identifying any. No licence, public domain basis or rights statement for any part of the corpus was located on 31 Aug 2026.

Sources are named, the licensing basis is stated, and the update cadence is published, which is the full set. Blue J identifies its content as primary authoritative tax material together with Tax Notes and IBFD, and the feature matrix breaks it down into case law, primary source materials and secondary source materials curated by tax experts. The licensing basis is not left to inference: section 13 of the Terms of Use reproduces Tax Analysts' own terms, referring throughout to Licensed Content and the licensed materials, disclaiming warranties on the publisher's behalf, and committing the publisher to defend infringement claims, which is what a content licence looks like on the page. Cadence is stated twice, with the database of primary tax sources described as current and updated daily. Two limits belong on the record: no equivalent licensing statement was located for IBFD, which is named but whose basis is not set out, and no coverage dates or corpus size are published on any first-party surface.

The platform operates its own citator and says how it is built. BCITE is described as Bloomberg Law's court opinion citator, using machine learning and natural language processing to let a user determine quickly whether the holding of an opinion is still valid law. That is a first-party treatment signal with a stated method rather than a licensed commercial citator or a prompt telling the reader to go and check. The description is where the disclosure ends: no coverage statement for which courts or date ranges BCITE treats, no accuracy or error rate, no explanation of how a treatment determination is reached beyond naming the techniques, and nothing published on whether the generative features surface BCITE status alongside authority they cite. Read 31 Aug 2026.

Checked the home page, how it works, the pricing page and its full feature matrix, and the Terms of Use on 2 September 2026. No public material addresses whether an authority the product returns is still good law. The nearest claims concern the freshness of the collection rather than the standing of any individual authority: the database of primary tax sources is described as current and updated daily, and the FAQ answers the recency question in those terms. Neither speaks to superseded regulations, revoked rulings, or decisions overtaken on appeal, which is the question this signal records. It bites here because the product retrieves case law, Treasury regulations and administrative rulings and presents them as the basis for a defensible answer.

An abstention path is documented with its triggers named. The vendor states that the Bloomberg Law Answers and AI Assistant features will abstain from answering when there is no available content covering the topic in question, or when the query falls outside legal scope. It sets that alongside an unusually direct acknowledgement that AI-generated responses are never fully trustworthy and carry a hallucination risk, and advice that customers validate responses against primary sources. What is not published is any demonstration: no worked example of a refusal, no evaluation or measurement of how often the behaviour fires or fails, and no confidence or grounding score surfaced to the user. Documented rather than demonstrable on that basis, checked 31 Aug 2026.

Checked the home page, how it works, the pricing page and feature matrix, the security page, the Subscriber Agreement and the Terms of Use on 2 September 2026. No current material describes what the product does when it cannot ground an answer, and no confidence or grounding indicator is described. One statement of abstention exists but not on any current surface: a company announcement issued in August 2023 quotes Blue J's head of legal research saying the platform is trained to admit when it is not able to formulate an answer supported by authoritative sources. That is a claim in a three-year-old press release rather than documented behaviour, and nothing on the site today repeats or elaborates it, so it is recorded here for a later grader rather than relied on. The current surfaces answer the adjacent question instead, offering the user tools to verify what was produced.

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 publisher's name, alongside 2026 sanctions trackers, law library commentary and trade press. 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. Two appearances should be distinguished from a hit, since the name recurs constantly in this literature: practitioner guidance repeatedly names Bloomberg Law as one of the databases against which a citation should be verified, and the vendor's own newsroom covers the sanctions phenomenon as a reporting subject.

Searched the AI Hallucination Cases database maintained by Damien Charlotin, and reporting drawing on it, on 2 September 2026 on both the product name Blue J and the corporate names Blue J Legal Inc. and BJL US Inc. No court order, opinion or disciplinary record naming the product was located. This is a statement about the public record rather than a finding about the product. One structural point: tax practice generates filings before the Tax Court and administrative submissions to revenue authorities rather than the general litigation filings that dominate the database, and the buyer base is largely accounting firms whose work product is often not filed by them at all, so the exposure this signal tracks is shaped differently here than for a litigation research tool.

Searched the approach-to-AI page, the AI product page, the about page, the pricing page, both agreements, the privacy policy and the trust centre landing page on 31 Aug 2026. No bar or ethics authority is engaged with anywhere in the vendor's product or policy material, including ABA Formal Opinion 512 and any state bar guidance, and nothing maps the product to a professional conduct obligation a firm could hand to its risk committee. The advice-line disclaimers in both agreements and the validation guidance on the approach-to-AI page are the vendor stating its own position rather than addressing the rules its buyers are bound by, and neither cites the guidance those rules come from. One distinction worth drawing, because it cuts the other way superficially: Bloomberg Law's newsroom covers legal ethics and AI sanctions extensively as an editorial subject, which is journalism about the guidance rather than the product being measured against it.

Checked the home page, how it works, the pricing page, the security page, the Subscriber Agreement, the Terms of Use and the privacy notice on 2 September 2026. No public material engages with professional guidance from any body governing this product's users. Neither ABA Formal Opinion 512 nor any state bar opinion is named, and no AICPA or state accountancy board guidance appears either, which matters because the named customer base is predominantly CPA firms whose AI use is governed by that guidance rather than by bar rules. Blue J does address the underlying professional obligation in substance, at section 7 of the Subscriber Agreement and section 12 of the Terms of Use, but as its own framing rather than by reference to the standards its users are bound by.

Efficiency claims without a word on the client's side of the bill. The vendor's material is built around time and cost saved: AI applications that help customers maximise efficiency and complete research faster, summarisation to speed consumption of news and research, and a flat all-inclusive subscription pitched on the basis that occasional use of a tool costs nothing extra. A customer quote on the pricing page goes further, describing reduced annual legal fees from fewer calls to outside counsel. Searched the pricing page, the AI product page, the approach-to-AI page and both agreements 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 an AI answer informs advice. The flat pricing model does remove the disbursement question that metered research platforms create, which is adjacent to this signal without meeting it.

Savings are claimed, quantified, and converted into money without the billing question ever being reached. The home page publishes three hours saved per user per week and 75 per cent less time spent on research, and the pricing page carries a calculator that turns a firm's headcount and research hours into recovered capacity, illustrated at 208 hours, 412,000 dollars of billable capacity and eighteen additional client matters annually, with the basis given only as average savings observed among Blue J users. Expressing the benefit in billable dollars puts the question squarely in view and nothing answers it: no published material addresses how AI-assisted work should be recorded, billed or disclosed to a client, and no per matter record of AI-assisted work was located. The point has force here because the buyer is a professional firm that bills clients for the time the tool compresses.

The artifacts a firm would need are catalogued on the trust centre and their contents could not be established from outside. The portal, linked from every product page footer and reachable without a sales conversation, lists a subprocessors entry, a data processing agreement, a data subject requests entry, a security whitepaper, a SIG Lite self-assessment and a document titled AI Frequently Asked Questions. None of those rendered its contents on fetch on 31 Aug 2026, no subprocessor or model provider is named anywhere on the portal landing page or on any Bloomberg Law page, and the document library is split into public and private with a get-access request whose tier the portal does not state. Recorded at the lower tier on that basis: the material demonstrably exists and is catalogued, and whether a firm can obtain it without an executed agreement is not stated. That is a better position than nothing published and short of a list a firm could forward to its client today.

A current subprocessor list is published openly on the security page and it is the most complete located in this pull: thirty-one named legal entities with a description of each role and its location, covering the AI layer explicitly with OpenAI for large language model AI, Google for LLM and AI services, Microsoft for cloud AI services, Pinecone for vector database services and ValsAI for benchmark testing, alongside infrastructure and business tooling, and including both Blue J affiliates as subprocessors in their own right. Section 11 of the Terms of Use ties the list into the contract, taking the customer's general authorisation for the subprocessors listed at that address. The reason this is not the top value is the third limb: no forwardable client-facing disclosure pack exists, the SOC 2 report requires a signed non-disclosure agreement, and no data processing agreement is published, so the material a firm would hand to a client is available only in part.

Sources are traceable per answer and the AI provenance a standing order asks for is not recorded. The vendor commits that when generative AI provides an answer or suggests text it includes links to source material or an explanation of why it returned that result, that Bloomberg Law Answers provides citations and links to the supporting authorities used in producing an answer, and that inline citations offer one-click verification to source, so a lawyer can produce the authorities behind an answer and check each one. It also commits to documenting where and when its products employ generative AI, which is the disclosure question at the platform level. What cannot be established: no model is identified or versioned anywhere, so which system produced a given passage is unknowable; nothing records who verified an answer; and no export designed for an AI-use disclosure or certification was located on 31 Aug 2026.

Checked the home page, how it works, the pricing page and feature matrix, the security page, the Subscriber Agreement and the Terms of Use on 2 September 2026. Nothing addresses standing orders, AI use disclosure or any certification that citations were checked by a person. The product does produce an artifact a professional could keep, since answers carry inline citations and source lists and can be exported to PDF or Word, but nothing describes that export as a record of model use, and it does not capture which model produced which passage, what was retrieved, or who reviewed it. The obligation this signal tracks also lands differently in tax practice, where work more often reaches a revenue authority than a court, and the vendor engages neither setting.