IPRally vs PatSnap: how they compare in 2026

The machine learning is the mechanism the buyer pays for. IPRally's proprietary Graph AI, described on the features page as trained by millions of patent examiner citations, is what executes the search: a user enters a free-text description, a publication number, an image or a hand-drawn search graph, and the model returns ranked prior art without Boolean construction. The company's own positioning is that this replaces query building rather than accelerating it, and the Agent page states the case directly against alternatives, that generic large language models cannot search the patent index and that an agent is only as good as the search technology underneath it. Remove the models and what remains is a full-text patent database with the Boolean search and filters IPRally ships as a secondary, industry-standard option, which is not the product anyone buys. This is the AI-native counterpart to Alt Legal and Anaqua in the same lane, both of which graded C, and the contrast is the axis discriminating rather than an unusually generous read.

A data platform that existed for over a decade before the models, with a substantial AI layer built on top of it. Founded 2007, and the asset underneath is the corpus: more than 2 billion structured data points and over 210 million patent, science and technology records with search, landscaping, dashboards and analytics that predate generative AI and function without it. Eureka is described by the vendor itself as a capability layer rather than a separate platform, which is an accurate self assessment and is credited as such. The AI layer is genuinely substantial rather than decorative, spanning domain specific agents for prior art, freedom to operate, drafting, translation and life sciences work, and the vendor describes a four stage data pipeline built to feed it. Graded B on the same basis as Relativity, Everlaw and Lexis+ AI: a mature platform hosts the model layer rather than depending on it. Remove the models and PatSnap remains a working patent analytics business. This is the first record in ip-and-patents below A on this axis, and the reason is age rather than weakness.

Grounding is real, documented and inspectable, and the measurement stops short of a test an outsider could run. The Agent page states that every search case can be opened and inspected, that AI reasoning and citations are given at the feature level, and that every claim assessment, feature mapping and relevance score is traceable to a source document in IPRally's patent index. The retrieval method is described rather than asserted: semantic search over graph representations, patent families, and following the examiner citation trail outward from the strongest documents. One measured figure is published, a 14 per cent improvement in search recall against a full patent text single search, but the comparison is to IPRally's own baseline, no test set is described, and no failure mode is named. The Generative AI Additional Terms of 14 March 2025 disclaim any warranty that outputs are correct and recommend checking them against a primary source, which is an honest position rather than an accuracy claim. Two published corpus figures also disagree, 120 million patents on the pricing page against 130 million on the Agent page. The citator limb does not bite here, since the product retrieves patent documents rather than legal authority whose subsequent history could be checked.

THE FIRST PUBLISHED BENCHMARK WITH A STATED METHODOLOGY AND A NAMED COMPARATOR IN 51 RECORDS. PatentBench is published as a benchmark built for real patent work, stated to measure specific tasks against structured data, legal context and source linked evidence, with a full methodology the vendor links rather than summarises. Published results: a 77 percent hit rate on design freedom to operate, and a claim of being 83 times more accurate than ChatGPT on that measure. Naming the comparator and the task is what separates this from the unfalsifiable claims elsewhere on this index, because a reader can locate the methodology and disagree with it. Grounding is described architecturally as well: retrieval augmented generation over a governed corpus, a four stage pipeline the vendor states reduces hallucinations through governance at each stage, and source linked evidence as a stated property of the benchmark. Graded A because this axis rewards verifiable evidence checkable by an outsider without contacting the vendor, and a published methodology with a stated result meets it. Held short of perfection and recorded plainly: the benchmark is the vendor's own rather than independent, the 83 times figure is a ratio against a general purpose chatbot on a task it was never built for, which flatters the comparison, and the methodology page itself was not read in this pass, making this a correction candidate in both directions.

The control structure is unusually well described for this corpus and two limbs are missing. What runs unattended is stated: Agent analyses an uploaded disclosure, identifies claims and features, searches and builds the feature chart. Where it stops is also stated, since the page says Agent verifies its findings with the user before searching, and the depth is a user setting described as quick triage or litigation-grade digging. The review surface is concrete, with every search case openable, reasoning and citations at feature level, and output handing off into a full IPRally project so a professional continues the analysis, framed as designed for professional review, not blind trust. The Generative AI Additional Terms add a contractual layer of control: generative features are always marked with an identifier before use, an administrator enables them for the organisation, and the customer can disable them at any time. What is not published is any statement of what happens after the system is wrong, and no abstention path is described for Ask AI, which is why this sits below the top band rather than at it.

Agent language is used extensively and no oversight model is published. The platform is described as agentic across IP and life sciences, with named agents including an FTO Agent, a Lead Compound analyser, an Antibody Target Predictor and a Markush Claim Drafting agent, and a published customer account describes completing infringement analysis for new products in hours rather than weeks. Data governance is described at each stage of the pipeline, which is governance of inputs rather than of agent behaviour. Nothing published states what an agent may complete without human review, whether any output can be relied on without verification, what confidence signal accompanies an agent result, or where a human checkpoint sits in an agentic workflow that runs from query to report. The gap matters because the outputs described are decision grade: a freedom to operate assessment that clears a product for launch, and a novelty judgement that decides whether a molecule is pursued. Checked the Eureka product pages, the life sciences landing page, the home page and the Open Platform material on 29 Aug 2026.

There is a named customer with a figure, and no date and no method behind it. The TLB case study identifies TLB GmbH, the central technology transfer office for universities in Baden-Wurttemberg, quotes Dr Frank Schlotter, Authorized Officer, by name and title, and reports an approximate 50 per cent reduction in time spent on patent searches, with the specific features credited, Smart search for query preparation and Ask AI during review. It is downloadable as a PDF and describes what the organisation did before, extensive keyword planning and manual Boolean queries plus a different AI tool they were not satisfied with. Five further case studies are published under named organisations including Unilever, RPX Corporation, Ossur, Metsa Group and Perl IP Consulting, and a testimonial on the Agent page names Josh Walling, Senior Patent Agent at Milwaukee Tool. What holds this below the top band is that no case study carries a date, and the headline figure is the customer's own estimate of time saved rather than a measurement a reader could assess. Only the TLB study was opened on 1 September 2026.

Named customers of the highest recognisability, a named executive, and independent third party presence, all dated and checkable. Named customers include NASA, Tesla, Vodafone, MIT and General Electric, reported in independent trade coverage rather than only on vendor surfaces, plus NCH Corporation with a named executive, Michael Schuster, VP of Product Development and Innovation, quoted by name and title on vendor material. Corporate history is a matter of public record: founded 2007, unicorn valuation reached in 2021 following a $300m Series E led by SoftBank Vision Fund 2 and Tencent, with Jeffrey Tiong named as chief executive. Independent evidence extends beyond a press release: a Gartner Peer Insights listing, a SoftwareOne marketplace listing, and a hands on product review by a named trade publication that obtained a trial and published its assessment. An outsider can verify the company, the customers, the funding and the product without contacting the vendor. Held short of perfection because outcome claims carry no methodology, being productivity boosted by up to 60 percent and workflows 90 percent faster, and one customer quote describing hundreds of thousands of dollars saved is not attributed to a named organisation.

The commitments a buyer most wants are in the agreement, and the segregation question is unanswered. The Terms of Service of 5 April 2024 state that IPRally will not gain ownership of user-generated content, will not share, distribute or disclose it to any third party, and will not use it for any purpose other than performing the functions of the Service, and give the customer a right to permanently delete all of it. The Generative AI Additional Terms go further than anything else located in this pull on the third-party model question, requiring that any external provider must not use input or output to train its models and must not claim ownership, and stating that user-generated content is never sent outside IPRally's own cloud environment unless the customer specifically approves it. Against that, privilege and work product are never mentioned on any surface read, and nothing addresses segregation between users or between matters inside a customer account beyond company-specific encryption keys and user-level feature rights. For an in-house IP department that is close to sufficient under the band as amended, but IPRally also sells to law firms, and for that segment the matter-level question is not answered.

General enterprise confidentiality is evidenced and the legal professional dimension is absent, and the buyer profile explains why without excusing it. Published: strict data isolation, ISO certified security policies, encryption in transit and at rest, consent management, GDPR and CCPA compliance, and a positioning as a privacy first partner for enterprises. Those are real controls covering the confidential research and portfolio material customers upload. What does not appear anywhere: attorney client privilege, work product, or any acknowledgement that a law firm user's material carries professional obligations distinct from a corporate research team's. The distinction from category peers is instructive rather than incidental. Patlytics and Solve Intelligence sell primarily to patent practitioners and both reached A here by engaging the professional rules; PatSnap sells across research and development, strategy, life sciences and legal, and addresses confidentiality as an enterprise data question throughout. That is coherent for its market and it leaves the practitioner question unanswered. Checked the trust centre material, the home page, the Eureka pages and the security statements on 29 Aug 2026.

What is published is a position on AI output reliability, not a position on the professional responsibility line, and the two are being counted separately here. The Generative AI Additional Terms tell users they are solely responsible for content generated, that they should independently verify it, and that they assume the entire risk, and the Agent page repeats that Agent does the searching while the user makes the decisions. Nothing located states that IPRally is not a law firm or does not provide legal advice, names a jurisdiction limit, addresses a practitioner's supervision or competence duties, or engages with ABA Formal Opinion 512, USPTO practitioner guidance or any EPO or national equivalent. This matters more than usual because the Agent page markets a should-we-file signal to R&D and innovation teams explicitly described as needing no patent expertise, which puts a patentability assessment in front of non-practitioners with only a general verification disclaimer behind it. Searched the home page, the features page, the Agent page, the pricing page, the legal hub carrying all four agreements, the about page and the PR page on 1 September 2026.

Not located. The platform generates freedom to operate assessments, novelty and design patent risk analyses, and drafts patent claims including Markush structures, all of which are outputs a practitioner signs off or a business relies on, and the customer accounts published describe non lawyers in research and development using them for risk management decisions. Nothing published addresses whether output constitutes legal advice, what the role of qualified counsel is where an engineer or a chemist runs a freedom to operate agent directly, or any professional conduct framework. The question is arguably more live for this vendor than for its category peers precisely because its named buyers include scientists and engineers rather than only attorneys. Checked the home page, the Eureka pages, the life sciences landing page, the trust centre material and the site navigation on 29 Aug 2026.

Principles are published and no mechanism sits behind them. The about page states three company values, transparency defined as clear, explainable AI solutions so a user always knows how decisions are made, simplicity and reliability, and the features page repeats explainability as a product property. That is an AI-specific principle rather than a generic value set, which keeps this off the bottom band. What is absent is everything that would make it auditable: no responsible AI or AI governance page exists, no ISO 42001 or EU AI Act position was located, nobody inside IPRally is named as accountable for model behaviour, no pre-release testing regime is described, and no evaluation of uneven output has been published for a system that classifies technology and scores relevance. Searched the home page, features, about, the legal hub and the help centre collection list on 1 September 2026. The identifier marking and customer opt-out for generative features are real governance mechanisms but they are spent on the oversight row and are not counted again here.

Data governance is described in operational detail and model governance is not. The vendor publishes that a four stage pipeline processes over 2 billion data points across 20 specialised domains with rigorous governance at every stage, stating this is what reduces hallucinations and improves output reliability, and names its retrieval architecture. PatentBench, graded on the Citation Accuracy axis, is a published evaluation instrument and its existence is evidence that measurement happens. What is absent is governance of the models themselves: no AI policy, no model card, no bias or fairness testing, no accuracy monitoring beyond the benchmark, no drift statement, no named governance body, no ISO 42001 and no EU AI Act positioning. The distinction being drawn is between governing the corpus and governing the system, and this vendor documents the first thoroughly and the second not at all. Compare Patlytics at B with an ISO 42001 certificate, and Solve Intelligence at B with a published continuous evaluation process and an EU AI Act self classification.

Every limb this band names is published and specific. Retention is stated with a number in the privacy policy updated 12 December 2025, which keeps personal data for a maximum of twelve months plus an ordinary backup period after the reason for holding it ends. Deletion is contractual on both sides, with the Terms of Service giving the customer a right to permanently delete all user-generated content and the Data Processing Agreement requiring deletion or return of personal data on termination. Access control is described as restricted to authorised personnel bound by confidentiality obligations, with SSO, optional adaptive multi-factor authentication, user-level feature rights and company-specific encryption keys. The subprocessor list published as DPA Annex B names nine processors, Okta, Google Cloud EMEA, HubSpot, Intercom, Planhat, Mailgun, Zapier, Slack and Pendo, each with its address, purpose, location of processing and the data categories involved. Incident practice is covered in DPA clause 9, requiring notice without undue delay and specifying the four contents of that notice, alongside a published Vulnerability Disclosure Policy last modified 17 February 2026 and a public system status page. The one soft edge is that the retention position is framed around personal data and no separate window is given for patent content or AI inputs; the subprocessor list is dated January 2025, which is noted rather than graded.

Isolation and compliance are stated and the training question is unanswered. Published: strict data isolation, ISO certified security policies, encryption in transit and at rest with named protocol and cipher, GDPR and CCPA compliance with stated consent management, and a self description as privacy first. Consent management is a genuine element and few records on this index name it. What was not located is any statement on whether customer uploaded documents, search histories, proprietary formulations or research materials are used to train or improve the domain models, any retention period, or any deletion right. The gap is material for this product because customers upload their own novel compounds, formulations and unpublished technical material into an analysis platform, which is competitively sensitive in a way patent search over public data is not. Compare Solve Intelligence at A and DeepIP at B in this category, both of which state a retention or training position plainly. Checked the trust centre material, the home page, the Eureka pages and the security statements on 29 Aug 2026.

The allocation of loss is published, readable before signing, and runs entirely against the buyer. The Terms of Service cap liability at the amount paid in the twelve months before the event, and state expressly that IPRally is not responsible for damage caused by either correct or potentially incorrect data in the Service, which is the exposure this product actually creates. The Generative AI Additional Terms reinforce it, providing outputs as is with no warranty of correctness, placing the entire risk on the user, and disclaiming liability for the output of external model services and for security failures at third-party providers. The Data Processing Agreement carries the same cap and does name carve-outs, excluding wilful misconduct and gross negligence from the limitation. No indemnity of any kind runs to the customer, no warranty attaches to output, and no insurance position is stated. This is a C rather than a D because a buyer can read the whole allocation before signing and the carve-outs are named, which is more than the two ip-and-patents vendors already in the index offer, neither of which publishes a customer agreement at all.

No published position located. Nothing was found on liability for AI output, warranty, service levels or remedy. The exposure here is unusual on this index because the consequential outputs are business decisions rather than filings: a published customer account describes the FTO Agent transforming risk management by completing infringement analysis in hours, and another describes lead compound screening acting as the first gate in a research and development process, so a false clearance sends a product to market and a false negative on novelty kills a viable molecule. The vendor publishes a 77 percent hit rate on design freedom to operate, which by its own arithmetic means roughly one in four is missed, and no published position addresses who carries that. Checked the home page, the Eureka pages, the life sciences landing page, the trust centre material and the site navigation on 29 Aug 2026.

The product largely stands alone and the connective tissue that exists is thin. What is documented is export rather than integration: custom exporting to Excel, PDF and Word, sharing of searches, collections and monitorings inside the account, single sign-on as an optional module, and a reference on the features page to unspecified tool interoperability. No connection to any IP management or docketing system, document management system, or patent office filing system was located, and there is no integrations page anywhere in the site navigation. A first-party blog post dated 23 July 2026 describes a new IPRally API as shipped and MCP support as upcoming, and the future-tense half of that is not evidence of anything today. The help centre carries a collection titled product news, data coverage, interoperability, holding three articles, which was not opened on 1 September 2026, so this grade is rebuttable if that collection contains a usable integration reference. Checked the home page, features, pricing, the full product navigation, the legal hub and the help centre collection list.

THE MOST COMPLETE INTEGRATION DISCLOSURE ON THE INDEX. The Patsnap Open Platform is a dedicated developer product, not a connector list, offering four distinct integration surfaces: REST APIs, MCP servers, UI widgets and agent skills, with a published Developer Center and documentation. Coverage is enumerated rather than claimed: more than 20 patent APIs, more than 25 corporate APIs and more than 16 infringement APIs. The MCP server offering is the notable element, since it exposes the corpus and its agents as tools to external AI systems, and the vendor names the frameworks it targets including LangChain, AutoGen and function calling workflows, which is integration aimed at how AI systems are actually assembled in 2026. Agent skills as a published surface goes further still. A buyer can obtain an API key from the site. Graded A because a developer can assess, obtain and build against this without contacting sales, which is the top of this axis, and because the surfaces are enumerated with counts rather than asserted. Held short of perfection because no named IP management or docketing system integration was located, which is the one thing a prosecution practice would ask for and which DeepIP states.

Residency is answered plainly and the tenancy model is not. The privacy policy states that the servers directly used by IPRally are located in the European Union, and the subprocessor annex names Google Cloud EMEA Limited of Dublin as the main cloud provider with location of processing given as the EU, and gives a processing location for each of the other eight subprocessors individually. The Data Processing Agreement clause 4 prohibits transfer of personal data outside the EU or EEA without the customer's prior written authorisation and requires standard contractual clauses where a transfer occurs. Unusually, the processing question is answered for the AI specifically: the Generative AI Additional Terms distinguish In-Cloud Services running inside IPRally's own cloud environment from External Services reached by API outside it, state that core models are hosted by IPRally itself, and commit that user-generated content is not sent outside that environment without customer approval. That is the precise limb that held Anaqua at B, answered here. What is missing is any statement of the tenancy model, single or multi-tenant, and there is no choice of region, since the EU is the only location offered.

Nothing located. No hosting provider is named, no region or data residency commitment is published, and no single tenant, dedicated or on premise option is described. The absence is more consequential for this vendor than for most on the index: the company is headquartered in Singapore with named investors including SoftBank and Tencent and serves customers including NASA, MIT and General Electric, so the jurisdiction in which research material and unpublished technical data is processed is a question those customers' own compliance functions would ask directly, and it is unanswered in public material. Stated GDPR and CCPA compliance establishes that regimes are addressed and says nothing about where data sits. Compare DeepIP at A for on premise deployment and Solve Intelligence at A for customer selectable jurisdiction. Checked the trust centre material, the home page, the Eureka pages, the Open Platform material and the site navigation on 29 Aug 2026.

The certification is real and stated twice on the property, on the home page and the features page, as ISO 27001 certification of the information security management system. What is not published anywhere readable is the scope of that certification, the certificate date or period, or the name of the auditor, so a buyer in 2026 cannot confirm the attestation is current. A trust centre exists and is linked from both the home page and the site footer at trust.iprally.com; it is hosted on Vanta, which its own page metadata confirms, and it returned a page frame with no body across the attempts made on 1 September 2026, with a search built on the portal's own terms failing to surface its contents. That block is a retrieval limit and nothing is graded against the vendor for it: this row rests on what is published on the readable surfaces, where a named standard appears without scope, date or a route to the report. It is what the portal might additionally offer that could not be confirmed, so B is a floor here rather than a ceiling.

A published trust centre with named standards and annual audits, held below the top by the type of the SOC report. Published on a dedicated trust centre: certification by the AICPA, ISO 27001:2022 with a statement that systems undergo rigorous annual external audits, encryption using TLS 1.3 and AES-256 in transit and at rest, GDPR and CCPA compliance, and consent management. Naming the ISO revision year and committing to annual external audit addresses currency, which most records on this index omit. THE LIMITING FACT: the SOC attestation is announced and described as SOC 2 Type 1. Type 1 assesses the design of controls at a single point in time; Type 2 tests whether those controls operated effectively across a period. Every other record on this index that reaches A on this axis holds a Type 2 or an equivalent, and the difference is not cosmetic. Held at B on that basis, alongside the absence of a named auditing firm and any certificate date for the ISO. Calibration within this category: Patlytics reaches A on three certifications with named penetration testing partners, and this record publishes a trust centre with a weaker attestation type.

This is the most complete model supply chain disclosure located in the pull, and it sits in a contract rather than on a marketing page. The Generative AI Additional Terms of 14 March 2025 name the third-party providers IPRally may use, Anthropic PBC and OpenAI LLC as External Services reached through their APIs, and Google LLC as an In-Cloud Service, specifying Gemini models and all models in Model Garden meeting stated requirements, with links to each provider's terms. Where they run is stated and the distinction is defined: External Services host models outside IPRally's cloud environment, In-Cloud Services run within it, and the features page adds that IPRally hosts its core AI models itself. The document also sets the minimum requirements a provider must meet to be used at all, no training on input or output and no ownership claim over either. The change route is the limb a strict reader may want to test: providers may be added or removed at IPRally's sole discretion, but the list itself sits inside a published agreement whose amendments are posted, with material changes notified by email or at login under the Terms of Service, so a customer can see the list move rather than being told separately that it has.

Nothing located. No foundation model provider, model family or version is named, and no subprocessor list was found. The vendor describes its architecture in technical terms, naming retrieval augmented generation and what it calls retrieval augmented thinking, and describes domain specific and domain trained models, all of which characterise method rather than provenance. Independent product review notes an evolving AI backbone, which implies underlying models change over time and makes identification more rather than less relevant for a buyer assessing consistency. For a platform whose Open Platform exposes agents to external systems through MCP servers, the identity of the models behind those agents is a question a developer integrating them would ask. Checked the home page, the Eureka pages, the Open Platform material, the trust centre material and independent review material on 29 Aug 2026.

The unit and the structure are published without the figure, which is the same shape graded B on Juro and SpotDraft. Two plans are named, Individual for a single user and Team for two to unlimited users, so the unit of charge is explicit, and the comparison table below them is the most detailed in this pull, running to roughly sixty line items across search, Boolean, review, AI assistants, monitoring, collections, collaboration, classification, coverage, security and support, each marked as included, optional or coming soon. Which components are chargeable extras is named rather than implied, covering custom taxonomy AI classifiers, AI classification of monitoring results, simplified R&D licences, SSO and adaptive multi-factor authentication. A three-day free trial is self-serve. No number appears anywhere: there is no rate, no band, no floor and no currency, the Team plan routes to Contact Sales, and the Agent product is described separately as value-based and agreed per customer. This sits at the lower end of the band, since unlike Juro no volume bands or currencies are given and nothing states what implementation adds.

The pricing structure is disclosed and no figure is. Published through an independent analyst platform and consistent with vendor material: a subscription model with multiple tiers varying by features, user numbers and access levels, with details provided on request based on customer requirements, and a named Enterprise plan that includes specified agents such as the Antibody Target Predictor and Markush Claim Drafting. Knowing that the meter runs on seats and feature tier, and knowing which capabilities sit behind the top tier, tells a buyer how the bill will grow and which plan they would need, which is more structural disclosure than most of this index provides. Two self serve entry points exist and are credited as reducing the barrier: a free trial of Eureka and an API key obtainable from the Open Platform. Held at C because no price, range or entry point figure is published at any level and every commercial route ends in a quote request. Source basis recorded as Third Party Estimated because the tier structure detail comes from an independent analyst listing rather than a vendor pricing page, which was not located.

Coverage is described with real substance and the outer boundary is left open. Five role-based solution pages address the patent searcher, patent analyst, patent manager, patent attorney and head of IP, and the Agent page segments its audience into IP leaders, patent professionals, R&D and innovation teams, and law firms, with a paragraph explaining what each gets. Five use cases are documented as supported, novelty and patentability, invalidity, state of the art, freedom to operate and classification. Corpus coverage is quantified as 58 jurisdictions and 120 million or more patents with machine translations, and the customer roster demonstrates the range, from corporate IP departments including Unilever, Bosch and Nestle through a public authority in the Danish Patent Office to tech transfer offices such as TLB and firms such as Perl IP Consulting and Laine IP. Some limits are stated at feature level, with Agent supporting novelty and invalidity while freedom to operate is on the roadmap and Boolean monitoring marked coming soon. What is not stated is where the product stops as a matter of coverage, with no statement on team size, portfolio size, or the fact that the platform addresses patents rather than the wider IP estate.

The largest and most specifically characterised corpus in the category, stated on the dimensions a researcher checks. Scale: more than 2 billion structured data points and over 210 million global patent, science and technology records, spanning patents, scientific literature, litigation records, clinical data and technical knowledge across 20 specialised domains. Breadth beyond patents is the distinguishing element, since prior art is not confined to patents and this is the only record in the category to state coverage of scientific literature and clinical data alongside patent documents. Industry coverage is enumerated across agriculture and chemicals, consumer goods, food and drink, life sciences, automotive, oil and gas, professional services, aviation and aerospace and education. Domain depth is demonstrated rather than claimed in life sciences, with named capability for sequences, Markush structures, structure activity relationship extraction, clinical trial comparison and druggability. Patent translation trained on global patent data addresses cross jurisdiction work directly. Graded A because coverage is enumerated with counts, domains and named artifact types a buyer can test against their own work. Held short of perfection because no update lag or refresh frequency is published and no patent office list is given.

The prohibition sits in the agreements rather than on a policy page, and it runs in two directions. The Terms of Service of 5 April 2024 limit IPRally itself to using user-generated content only as required to perform the functions of the Service, and state it will not gain ownership of that content or disclose it to any third party, which forecloses training without using the word. The Generative AI Additional Terms of 14 March 2025 address third-party models expressly, setting it as a minimum requirement that any external or in-cloud provider does not use input content, prompts or output to train its machine learning models and does not claim ownership of either. The home page adds that IPRally's own AI is trained exclusively on existing patent data. The quote recorded is the provider-side requirement, which is the explicit training language; the IPRally-side commitment is the purpose limitation in the Terms.

Silent. The quoted phrase is the closest the vendor comes to addressing customer content in the AI context and it describes isolation and compliance rather than use. No statement in either direction was located on whether customer uploaded documents, proprietary formulations, search histories or research materials are used to train or improve the domain specific models. The gap is material because of what this platform ingests: customers upload novel compounds, unpublished technical material and internal documents for analysis, and the vendor separately markets domain trained models across 20 specialised domains, so the question of whether one feeds the other is directly raised by the product design. Recorded as silent, not as a negative commitment. Correction candidate: the trust centre was identified but not entered in this pass. Checked the Eureka pages, the home page, the life sciences landing page, the trust centre summary and the Open Platform material on 29 Aug 2026.

No retention period for prompts or outputs is published, and no zero-retention option is stated, but the customer holds a contractual deletion right over the material. The Terms of Service allow the customer to permanently delete all user-generated content, subject to ordinary backup storage periods, and the Generative AI Additional Terms define prompts as built from user input forming part of that same user-generated content, so the right reaches what a user typed into Ask AI or a search field. The privacy policy's twelve-month retention figure governs personal data rather than patent content or AI inputs. The control is a delete-on-demand right in the agreement rather than a configurable retention window in the product.

Not addressed. No retention period is published for uploaded documents, agent queries, generated reports or analysis outputs, and nothing indicates whether retention is configurable or whether a customer can require deletion. Consent management is named as a capability, which addresses the collection of personal data under GDPR and CCPA rather than the duration for which uploaded research material is held. Compare DeepIP at disclosed fixed with a stated zero retention policy, Solve Intelligence at customer configurable zero, and Patlytics at disclosed fixed with a 90 day window, all within this category. Checked the trust centre summary, the Eureka pages, the home page and the Open Platform material on 29 Aug 2026.

IPRally operates its own permission model rather than inheriting one, and the documented level is the user and the company rather than the matter. The pricing comparison names user-level feature rights management, single sign-on, optional adaptive multi-factor authentication and company-specific encryption keys, and the Terms of Service describe admin users setting licence counts and enabling or disabling features for some or all users. Sharing of searches, collections and monitorings is an explicit act rather than a default. The classic ethical wall question bites less here than on a document-based product, because retrieval runs over a public patent corpus rather than over the firm's own files; what is not addressed is segregation between matters or teams inside a single customer account.

Claimed and not documented. The vendor states strict data isolation alongside ISO certified security policies and consent management, so a separation model is asserted. Nothing documents it: no statement of whether isolation operates between customers, between users within an organisation, or between projects, and no description of how the boundary applies to the agent layer when an agent runs analysis across a corpus. The scenario this product raises is different from a law firm conflicts question and is equally live: the platform is sold across research and development, strategy and legal within the same enterprise, and to competing enterprises in the same technical field, and nothing published describes what prevents one customer's uploaded formulation or unpublished compound from informing anything surfaced to another. No integration with a customer permission system was located that would let the platform inherit an existing model.

The commitment is present but narrower than the question. Clause 3(A) of the Data Processing Agreement requires IPRally to process personal data only on the customer's written instructions unless compelled by law, and in that case to inform the customer of the legal requirement before processing, unless the law prohibits telling them. That is a notice-before-disclosure commitment in the standard GDPR form. Its scope is personal data under the DPA; no equivalent clause covering user-generated content or search history was located in the Terms of Service, and the privacy policy separately reserves the right to disclose personal data to authorities under court order or subpoena without stating a notice duty. No transparency report was located on 1 September 2026.

Not addressed. No government or law enforcement request clause, no commitment to notify a customer before producing their data, and no transparency report were located. The question carries particular weight for this vendor and the note records why plainly rather than implying anything: the company is headquartered in Singapore with named investors including SoftBank Vision Fund 2 and Tencent, and its named customers include NASA, MIT and General Electric, so the jurisdictions in which a request could be served and the parties who could receive one are a live diligence question for customers uploading unpublished technical material. Nothing published addresses it in any direction. Checked the trust centre summary, the home page, the Eureka pages and the site navigation on 29 Aug 2026.

The product does not retrieve case law; its corpus is patent literature and it is identified by scale and by jurisdiction rather than by licence. The pricing page states coverage of 58 jurisdictions and 120 million or more patents with machine translations of non-English documents and downloadable PDFs, while the Agent page gives 130 million or more for the same index, a discrepancy noted rather than resolved. The training signal behind the ranking is named specifically as millions of patent examiner citations. The Terms of Service name Google Patents and Espacenet as third-party resources the Service may link to. No licence, rights basis or update cadence for the underlying patent data is stated anywhere located on 1 September 2026.

Named with the most specificity in the category, and no licence basis stated. The corpus is enumerated rather than gestured at: more than 2 billion structured data points and over 210 million global patent, science and technology records, spanning patents, scientific literature, litigation records, clinical data and technical knowledge across 20 specialised domains, with a four stage processing pipeline described as applying governance at every stage. Coverage of scientific literature and clinical data alongside patents is the distinguishing element and no category peer states it. What is absent is the basis on which any of it is held. Patent documents are published by offices as a condition of grant, but scientific literature is substantially copyrighted and clinical trial data is licensed or regulated, and nothing states what rights support their inclusion, from which publishers or registries, under what terms. No update lag, refresh frequency or historical date range is published either, despite the pipeline governance claim. The larger and more heterogeneous the corpus, the more the licensing question matters, and this is the largest corpus in the category.

No citator applies and the row is recorded rather than skipped. The product retrieves and analyses patent documents rather than legal authority whose subsequent history could be checked, so there is no treatment signal to surface and no good-law question in the sense this signal asks. The nearest analogue is registration currency: the pricing comparison lists bibliographic data, legal statuses and full specifications among the review features, so a user can see the status of a patent, which is a record of the register rather than a judgement about whether an authority still stands. Searched the features page, the pricing comparison, the Agent page and the use-case pages on 1 September 2026.

Not addressed, and applicable in the patent specific form. The platform performs freedom to operate analysis, design patent risk assessment, infringement risk identification and novelty assessment, all of which turn on the legal status of the references surfaced: whether a patent is granted and in force, lapsed for unpaid fees, expired, amended in reexamination, or invalidated in post grant proceedings. A freedom to operate clearance that treats a lapsed patent as a live blocking right, or misses one that has been revived, is wrong in a way no amount of retrieval quality fixes. Nothing published states whether legal status is tracked, surfaced or incorporated into the risk scoring, and litigation records are named as part of the corpus without any statement that outcomes are linked back to the patents they concern. Fourth consecutive record in this category to leave this signal unanswered, and the pattern is now firm across every ip-and-patents vendor built.

A relevance signal is exposed and an abstention path is not. The pricing comparison names AI relevancy scores and Boolean relevancy scores as shipped features, and relevance-feedback search lets a user flag strong results to pull similar ones, so the product surfaces a graded confidence in each retrieved document. That is a retrieval relevance score rather than a measure of confidence in a generated answer. Nothing located describes what Ask AI or Multi-patent Ask AI does when it cannot support an answer from the documents, and no no-answer or low-certainty path is documented. The Generative AI Additional Terms instruct the user to verify outputs against a primary source, which places the duty on the reader rather than describing a behaviour of the system.

Not addressed as a behaviour, with an aggregate accuracy figure published instead. The vendor publishes a 77 percent hit rate on design freedom to operate through PatentBench, which is an honest headline number and is graded on the Citation Accuracy axis. It describes performance in aggregate and says nothing about the individual result in front of a user: no confidence indication accompanying an agent output, no flag on a low certainty match, no statement of whether an FTO Agent will report that it could not reach a conclusion rather than returning one, and no description of what a user sees when the corpus does not support an answer. The distinction matters most on precisely the number published, since a stated 77 percent hit rate means roughly one in four is missed and nothing tells the user which. Compare Patlytics at documented in this same category, on colour coded confidence indicators surfaced at the point of reading.

Searched the AI Hallucination Cases database maintained by Damien Charlotin at HEC Paris, together with 2026 sanctions trackers and trade coverage, on 1 September 2026, on both the company name and the product name. No court order, opinion or disciplinary record naming IPRally or IPRally Technologies was located. This is a statement about the public record rather than a finding about the product. The failure mode this signal tracks also fits the product poorly: IPRally retrieves real patent documents from an index and its generative features summarise and analyse those documents, rather than producing citations to legal authority for filing.

None located, with the instrument named. General web searches combining the vendor and product names with court, order, sanction, fabricated citation and patent terms returned nothing on 29 Aug 2026, and no named docket database, patent office record system or court record tracker was searched. Recorded as a statement about what this search found, not as a clearance. The exposure shape differs from the drafting products in this category: this platform generates analyses and reports supporting business decisions rather than documents filed at a patent office, so the analogous adverse finding would be a proceeding addressing reliance on a defective freedom to operate clearance, which would surface as commercial litigation rather than in a prosecution file wrapper.

No engagement with any professional responsibility guidance was located. Nothing addresses ABA Formal Opinion 512, USPTO guidance for practitioners on artificial intelligence, EPO or national patent attorney guidance, or any state bar opinion, and no ethics or professional responsibility page exists on the property. This is a European vendor selling to patent attorneys in multiple jurisdictions as well as to US firms, and none of the relevant regimes is named. Searched the home page, the features page, the Agent page, the five solutions pages listed in the navigation, the legal hub carrying all four agreements, the about page and the PR and media page on 1 September 2026.

Not addressed. No named ethics opinion, no USPTO Rules of Professional Conduct reference, no 37 CFR citation and no bar guidance of any jurisdiction was located. The absence is coherent with the vendor's market position rather than an oversight: PatSnap sells across research and development, strategy, life sciences and legal, and the majority of its named customers are corporations rather than law firms, so its material addresses enterprise buyers rather than practitioners. That explains the silence without curing it, since the platform still generates freedom to operate assessments and drafts claims, and legal is a named buyer. Third of four records in this category at this value, with Patlytics the only exception. Checked the home page, the Eureka pages, the life sciences landing page and the site navigation on 29 Aug 2026.

Time savings are the central marketing claim and the client side of that equation is never addressed. The TLB case study reports an approximate 50 per cent reduction in search time, the Agent page is built on the promise of minutes rather than weeks, and it positions directly against the cost of outsourced search, contrasting reports that cost thousands with a predictable subscription. No guidance accompanies any of it on how a firm or a tech transfer office should bill for AI-assisted search work or disclose it to the client paying for the assessment. No per-matter record of AI-assisted work is described that a firm could interrogate for billing purposes.

Savings claims only, and the published figures describe enterprise outcomes rather than firm labour, which is a category first. Standard efficiency claims: productivity boosted by up to 60 percent and workflows completed 90 percent faster. The distinctive one is a customer quote stating the platform has literally saved hundreds of thousands of dollars in the context of evaluating partnerships and acquisitions, which is a claim about avoided transaction cost rather than about hours, and a second describing infringement analysis moving from weeks to hours. None carries methodology, baseline or period, and the six figure saving is not attributed to a named organisation. Nothing appears on the client's side of the equation: no position on how AI assisted analysis should be recorded where a firm bills a client for freedom to operate work, and no exportable record showing what portion of an assessment was machine generated.

Both artifacts a client-side AI clause asks for are published, and neither requires an agreement to obtain. DPA Annex B lists nine named subprocessors with addresses, purposes, processing locations and data categories, dated 24 January 2025. Separately, the Generative AI Additional Terms name the model providers themselves, Anthropic PBC, OpenAI LLC and Google LLC, distinguish which run inside IPRally's cloud from which are reached externally, and state the minimum requirements each must satisfy, no training on input or output and no ownership claim. The same document is forwardable to a client as disclosure material: it describes exactly what leaves the environment on an Ask AI query, gives a worked example, and records that an administrator can disable generative features entirely. This is the highest value on this signal recorded in the pull to date and the reasoning is set out in the build log for comparison against Litify and Anaqua.

On request, through a published trust centre with named standards behind it. The trust centre is a defined destination and the disclosures a firm could cite are specific: AICPA certification, SOC 2 Type 1, ISO 27001:2022 with stated annual external audits, TLS 1.3 and AES-256 encryption, GDPR and CCPA compliance, strict data isolation and consent management. Naming the ISO revision year and the audit cadence is more than most records manage. Held at on request rather than higher because nothing is downloadable and the pack is incomplete: no subprocessor list, no named model provider, no data processing agreement, no auditor identified, no certificate dates, and no route to the SOC report itself was located, with the trust centre gate not entered in this pass. The SOC being Type 1 rather than Type 2 also limits what a firm can represent onward to a client, since Type 1 speaks to control design at a point in time rather than operating effectiveness over a period.

Some of the record exists and no document-level export of it is described. Generative outputs are always marked in the interface with a specific identifier under the Generative AI Additional Terms, so a user can tell which material was AI-produced, and the Agent page states that every search case can be opened and inspected with AI reasoning and citations attached at feature level, which covers what was retrieved and why. Relevance marking and commenting record who reviewed a document. What is not described is any export tying a particular passage to the model that produced it, since the identifier marks that generative AI was used rather than which of the three named providers ran, and no certification artifact is offered. The obligation fits prosecution and portfolio work imperfectly, but the row is recorded rather than skipped.

Partial record, with a stronger source limb than most and the same two gaps. Source linked evidence is named as a property of the PatentBench methodology and the platform is built on a governed corpus with a four stage pipeline, so an assertion in a generated report is traceable to a document in a way a reader can follow, and the Open Platform exposes the underlying data through APIs so a claim can be independently re-queried against the same corpus. That last property is unusual and is a genuine disclosure strength: a party can verify the underlying data without the vendor's cooperation. The familiar limbs are absent. Nothing indicates that output records which model or agent produced it, which matters more here because independent review describes an evolving AI backbone, so the system behind a report generated last year may not be the one running today. And no human verification record is captured, so a party relying on a freedom to operate assessment in a dispute cannot evidence that a qualified person reviewed it.