P
PatSeer
Patent professionals use PatSeer to retrieve global patent and scientific records, assess inventions, track portfolios and share findings. The platform combines Boolean retrieval with semantic ranking driven by its ReleSense engine. Analysts can submit an invention description, refine the returned patents using relevant examples and open patent text and claims for review. Semantic Mapping builds an editable claim chart by associating elements of an independent claim with disclosures in selected documents.
PatAssist, AI Summaries, AI Recommender and an AI Classifier for reusable taxonomies support generative and sorting work, and PatSeer Designs adds image similarity for industrial design research.
The database holds more than 185 million records from 110 patent authorities and 86 design authorities, with almost 150 million in full text, 73 million machine translations, patent families, legal status, standard essential patents and litigation data. It draws on patent offices and licensed sources including IFI CLAIMS. The World Intellectual Property Organization lists PatSeer as a source in its specialized patent information access program, and the company reports more than 10,000 users.
Explorer targets company wide research access. Premier, PremierPlus and ProX add deeper research and analytics, including shared projects, portfolio benchmarking and visualization. The service describes a flat fee subscription and a two week trial. Its generative AI terms bar both PatSeer and external providers from using customer inputs or outputs to train generative models.
Alternatives to PatSeer
Other products that do the same kind of work as PatSeer, and what each one does.
- PPatent BotsPatent Bots is a suite of tools for patent preparation and prosecution from Patent Bots, Inc., founded in 2018 by patent attorney Jeff O'Neill.
- PPowerPatentPowerPatent is generative AI drafting software for patent applications, sold to patent attorneys, agents and the inventors who work with them.
- CClerqClerq is a patent research and intelligence platform for patent attorneys, in house intellectual property teams and organizations that watch patent portfolios.
- AArcPrimeArcPrime is an AI patent platform and IP management system for in house patent teams, built to replace a legacy IPMS rather than sit beside one.
- BBlack Hills AIBlack Hills AI makes intellectual property software and provides docketing and paralegal support for patent and trademark practices.
- IIP AuthorIP Author is an AI patent platform for practitioners, covering drafting, prior art search, office action responses, claim drafting, invention management and patent classification.
Capability grades
All 15 axes, graded from public sources on the date shown. Hover a grade to see what the letter means on that axis.
AI Centrality
How much of the product is actually AI. Whether the machine learning is the mechanism the buyer is paying for or a feature layered onto conventional software, and whether the vendor is specific about which is which.
ReleSense enables semantic retrieval and related features on a platform that also includes Boolean retrieval, patent families, legal events, alerts, and collaboration. The contemporary workspace reranks records against a description or seed patent and makes recommendations based on relevance. Semantic Mapping aligns elements of an independent claim with passages in selected prior art. PatAssist and image similarity handle additional workflows.
The database, shared projects, and record monitoring can be used independently of generative AI. The models drive how users find and examine relevant documents while other parts of the subscription organize the underlying data.
Citation Accuracy and Hallucination Disclosure
Whether the vendor publishes measured accuracy on citations and assertions, grounds output to primary sources, and says plainly what its system does when it does not know. Legal has a documented public record of fabricated citations reaching filed briefs, so an untested claim of accuracy is not evidence.
Analysts can examine patent documents and trace claim chart passages to the selected prior art references. Nearly 150 million records carry full text, with 73 million machine translations, so the text behind a mapped passage sits in the same database. In an OLED case study, PatSeer wrote a sample invention description and manually rated element overlap in the top 10 retrieved results. Each result got full, partial or zero points, weighted by position, which produced 88.6% for that example.
The experiment measures how relevant the retrieved patents were. It does not measure the accuracy of PatAssist citations or how often generative output is unsupported, which the AI terms cover with a separate accuracy disclaimer.
Autonomy and Oversight Model
What the system decides on its own, what a lawyer must approve, and whether the vendor documents where the review point sits. A tool that drafts under review and a tool that files without one are different products and different risks.
Analysts choose the source documents for Semantic Mapping and inspect the disclosed passage behind each claim element. The AI Classifier runs in automatic or moderated modes, so a reviewer can confirm its labels before they apply. The generative terms let the customer enable the features for its organization, ask for them to be disabled and approve sending user content to external services, and they point users to primary sources when using output.
There is no confidence level at which the product stops generating analysis, and no exception flow that requires a patent professional to approve a conclusion.
Operational and Outcome Evidence
Named, dated evidence that the product works in production at real firms or legal departments. Case studies with figures and identified customers count. Unattributed testimonials and launch announcements do not.
A case study describes University College London research using PatSeer for patent analysis. The World Intellectual Property Organization lists the product as a contributing database in its specialized patent information access program, which serves patent offices and research institutions in developing economies. PatSeer reports a community of more than 10,000 users and a 4.7 out of 5 rating on G2, with client reviews on its site.
Its benchmark series tests retrieval against written invention descriptions, including a flexible OLED screen example with its rating method and result. No named legal team's use is tied to a measured change in cost, review time or errors.
Privilege and Confidentiality Posture
How client confidences are handled: attorney client privilege and work product treatment, segregation of one client matter from another, whether client data trains any model, and what the vendor commits to in writing rather than in marketing.
PatSeer and its external service providers may not use user content or outputs to train, develop or fine tune generative models, and the home page adds that customer documents, lookups and activity never train AI. Approval is required before user content goes to an external service. The trust materials describe segregated containers in the shared deployment and dedicated accounts in the enterprise deployment, and engineer access is role controlled.
Privileged communications and legal work product are not treated explicitly, no deletion deadline applies to AI inputs and outputs, and there is no technical wall between matters within one firm's subscription.
UPL and Professional Responsibility Posture
Whether the vendor is clear that it supplies a tool rather than legal advice, who its audience is, and how it addresses unauthorized practice of law, competence and supervision duties, and jurisdiction limits. ABA Formal Opinion 512 is the reference point. Where the advice line is not the duty a product raises, the axis is read through the nearest professional duty it does raise: judicial conduct rules and the reviewing duty for products sold only to courts, and the duty to bill for time actually spent for products that draft time entries.
The offering is presented as patent research and analysis software for professionals, including in law firms, companies, and research institutes. Users select patent documents, inspect the mapped disclosures, and can review suggested results before using them in their work. The generative terms disclaim guarantees for correctness and direct users to verify material against primary sources. Public materials put patent evaluation in the professional workflow.
They give no detailed supervision guidance by jurisdiction and do not define which parts of a patentability, invalidity, or freedom to operate conclusion belong exclusively to counsel.
AI Governance and Bias Disclosure
Published governance over model behavior: who owns it inside the vendor, what is tested before release, and what is disclosed about disparate output across matter types, parties, or populations.
An engineer or attorney can inspect the source documents behind a claim chart and review its mapping. PatSeer labels which functions are generative and runs a series of retrieval relevance exercises on technology topics and element overlap, and the AI Classifier emails prediction performance statistics to its users. The policies name a data protection officer and describe security review and access. No owner for AI governance, release approval process or method for testing uneven output across authorities, languages or technology fields is described, and the benchmark examples break down no generative errors by group.
AI Safety and Data Stewardship
Retention, deletion, access control, and what happens to prompts and documents after they are processed. Whether the vendor states its subprocessors and its incident practice, or leaves the buyer to assume.
The privacy policy describes least privilege access, permitted categories of subprocessors, customer administration tools for deletion and export, and breach notification without undue delay. The generative terms identify Azure OpenAI and AWS Bedrock and say they cannot train on user content. The trust documentation describes data segregation and encryption, including vector data in dedicated environments. The policy and terms provide no defined period for keeping prompts or generative outputs at PatSeer or its named AI hosts, and the provider names in the AI terms do not form a complete subprocessor register.
AI Liability and Recourse
What the vendor stands behind contractually when its output is wrong. Indemnities, caps, carve outs, insurance, and whether any of it is published or only reachable through a negotiated agreement.
Outputs of the generative features are supplied as is, with no warranty of correctness, and their use is at the customer's risk, while the customer owns generated output subject to the agreement. The base end user license is available from PatSeer on request, and the website terms expressly exclude the patent application itself. The AI addendum carries no indemnity, capped remedy or insurance provision for an erroneous output. Wider contractual liability is set by the licensing and ordering documents.
Practice Systems Integration Depth
How deeply the product reaches into the systems legal work already lives in: document management such as iManage and NetDocuments, Word and Outlook, contract lifecycle management, matter management, e-billing, and court filing systems.
A public web API entry point exists for connecting applications to PatSeer. The privacy policy describes administrator configuration of access and transmission for customer enabled integrations. The database also receives patent records through the IFI CLAIMS Direct API under a licensing arrangement, and PatSeer has described distributing access through ProQuest Dialog. Analysts can import patent identifiers and export selected records in spreadsheet, document, XML, and CSV forms.
Public materials give less detail about how a contemporary deployment synchronizes projects with a law firm document system or an intellectual property docketing system.
Deployment Model and Data Residency
Where the software runs and where the data sits. Multi tenant cloud, single tenant, private deployment, on premises, and whether region of residence is a published option or an enterprise conversation.
The GDPR declaration places the standard web application and associated personal information on Amazon Web Services in Virginia. The trust documentation describes a shared Kubernetes cluster with segregated containers and a dedicated option with its own AWS account and cluster. The AI terms name Azure OpenAI and AWS Bedrock for some generative capabilities and require customer approval before external processing. Inference regions for those providers are not stated, and no region choice is offered for either the standard or the dedicated deployment.
Security Certifications and Trust Center
Independent attestation a buyer can pull without a sales call: SOC 2, ISO 27001, penetration test summaries, a trust center with current reports and named scope rather than a badge image.
The security description names EY as the firm behind a SOC 2 Type II examination covering security, availability and confidentiality, and PatSeer also states ISO/IEC 27001:2022 certification and GDPR compliance. Data is encrypted with AES 256 at rest and TLS 1.3 in transit, and access is limited, logged and governed by strong authentication. The service deployment documentation adds encrypted storage, logging and access controls for sensitive data.
A trust center is linked, but neither an audit period nor a copy of the report is available from it, and no certificate shows the covered system boundaries, issue date and expiry.
Model Supply Chain Disclosure
Which models sit underneath, whose they are, where they run, and whether the vendor commits to telling customers when that changes. A legal buyer inherits every dependency it cannot see.
ReleSense, the base retrieval engine, is trained on patent and scientific literature. The generative AI terms separately name Azure OpenAI and Anthropic models hosted through AWS Bedrock, and they describe the external environments, the customer's approval and the restrictions on providers' use of user content. Users are prompted to accept revised terms before continuing to use the capabilities. Model versions are not named, and there is no promise of advance notice when a model or provider changes without a change to the terms. Inference regions are also not stated.
Commercial Transparency
Whether a buyer can learn what this costs without entering a sales process: published rates, the unit being charged, what sits behind an enterprise tier, and what implementation adds.
PatSeer describes a flat fee subscription and offers a two week trial without a contract or payment card. Explorer is sold as company wide licensing for research teams. The plan comparison shows the differences between Premier, the PremierPlus feature group, ProX, Explorer and Designs, with PatAssist, Semantic Mapping, the AI Classifier, visualization and litigation data separating parts of the range. Paid rates come on request.
Setup, migration and dedicated environment fees are not listed, and there is no rate schedule showing how price changes with users, departments or data selected.
Firm and Practice Coverage
Who the product is actually built for. AmLaw, midlaw, small firm and solo, in house departments, government and courts, and which practice areas are supported rather than merely claimed.
Law firms, corporate patent and research teams, universities, service providers, and research institutes receive distinct use case descriptions. Explorer targets company wide research users, while Premier and ProX serve patent practitioners and analytics teams. Industrial Designs supports firms working with design registrations and image similarity. The World Intellectual Property Organization lists PatSeer in a program for eligible patent offices and research institutions in developing economies.
The plan and data descriptions show which offerings cover worldwide patent retrieval, scientific records, or industrial designs, and identify project editions that omit access to patent database retrieval.
7 public documents
The public pages on file for PatSeer, with the recorded signals each one supports and the date it was last read. Open any of them and check the reading against the record.
-
Client Data in Training, Prompt and Output Retention, Refusal and Uncertainty Behavior and 2 more
Read Nov 19, 2026
-
Good Law Verification, Billing and Fee Posture
Read Nov 19, 2026
-
Primary Law Corpus Provenance
Read Nov 19, 2026
-
Fabricated Citation Record
Read Nov 19, 2026
-
patseer.com/claim-charts-in-a-click-smarter-patent-analysis-with-semantic-mapping-webinar-recording1 signal
Court Disclosure Support
Read Nov 19, 2026
-
patseer.com/privacy-policy1 signal
Third Party Request and Subpoena Notice
Read Nov 19, 2026
-
patseer.com/trust-and-security1 signal
Ethical Walls and Matter Segregation
Read Nov 19, 2026
No published figure
- PatSeer quotes a flat fee subscription. Explorer has company wide licensing. Premier, PremierPlus, ProX, Explorer and Designs each have feature descriptions on the plan comparison. A two week trial is available without a contract or payment card.
Paid rates are quoted for each edition. Explorer is described as company wide licensing.
Implementation: No setup, migration or dedicated hosting charge is listed.
Note: The plan comparison shows feature differences and a trial. Paid amounts require a quote.
Legal Signals
What each signal meansA signal records what public sources say on the date shown. It is not a grade and it is not a recommendation. Where a signal reads Not addressed, it means the index did not locate the material in public sources on that date, which is a statement about disclosure rather than about the product.
Client Data in Training
Can material a lawyer puts into this product be used to train a model?
The published terms prohibit training on customer content. Not a policy page, the agreement.
Section 3.3 of the generative AI terms bars PatSeer and its providers from using customer inputs or outputs to train, develop or fine tune generative models. The home page adds that customer documents, lookups and activity are never used to train AI.
Prompt and Output Retention
How long does the product keep what a lawyer typed, and can that be set to zero?
No located public material states how long prompts and outputs are retained.
Administrators can request or start deletion and export of customer data, and vector data sits in dedicated environments. Neither the generative AI terms nor the privacy policy sets a retention period for prompts and results at PatSeer or at its AI hosts.
Ethical Walls and Matter Segregation
Does retrieval respect the firm’s ethical walls, or can the model read across them?
Segregation is asserted in public materials with no published detail on how it is enforced.
The shared deployment uses segregated containers, and enterprise customers can take a dedicated AWS account and single tenant Kubernetes cluster. Shared projects let teams curate and annotate datasets together, but no technical wall between matters inside one firm's subscription is described.
Third Party Request and Subpoena Notice
If someone subpoenas the vendor for a firm’s data, does the firm hear about it first?
No located term or policy addresses third party requests for customer data.
The privacy policy commits to notifying customers of a personal data breach without undue delay. It makes no matching promise for a subpoena or other compulsory request for a customer's matter data.
Primary Law Corpus Provenance
Where does the law in this product come from, and does the vendor have the right to use it?
The vendor names its primary law sources and the license or public domain basis for each, with an update cadence.
IFI CLAIMS licenses its global patent data to Gridlogics through CLAIMS Direct. Records also come from 110 patent authorities and 86 design authorities and are linked to status, families, machine translations and litigation data.
Good Law Verification
Does the product tell you when the authority it just cited has been overruled?
The vendor computes and surfaces subsequent history itself, with the method described.
Patent records carry legal status from INPADOC and supplementary sources including the USPTO and India's InPASS, alongside litigation data on standard essential patents and nonpracticing entities. This is a status signal on patents rather than judicial citation treatment.
Refusal and Uncertainty Behavior
What does the product do when the answer is not in the corpus?
No located public material addresses what the product does when it cannot ground an answer.
PatSeer points users to primary sources for verification, and the AI Classifier's moderated mode holds labels for review. No condition is described under which PatAssist declines to answer when the documents do not support a response.
Fabricated Citation Record
Does a public court record exist addressing fabricated or hallucinated legal citations in output from this product?
No court order, opinion or disciplinary record addressing fabricated or hallucinated legal citations produced by this product has been located as of the date shown. This is a statement about the public record on that one subject, not a finding about the product, and this signal is not a litigation history.
The AI Hallucination Cases database maintained by Damien Charlotin lists no case involving fabricated citations generated by PatSeer. Its benchmarks concern how relevant the retrieved patent documents are, and its claim charts point to passages in those documents.
Bar Guidance Alignment
Has the vendor engaged in public with the ethics opinions its buyers are bound by?
No located public material engages with bar or ethics guidance.
The generative terms call for verifying output against primary sources and leave correctness with the user. No workflow is tied to named bar guidance or to USPTO guidance on AI use in patent practice.
Billing and Fee Posture
Does the vendor address what happens to the bill when the work takes an hour instead of six?
The product sits inside a lawyer to client fee relationship and no located public material addresses billing, fee or disclosure treatment, with no savings claim published either.
Law firms, corporate patent teams and research institutes buy flat fee subscriptions for access to the tools. Nothing in the product marks AI assisted work or AI charges for a legal client's bill.
Outside Counsel Guideline Readiness
Can a firm get this vendor through a client’s AI clause without a bespoke negotiation?
A current subprocessor or model provider list is published.
The generative AI terms name Azure OpenAI and Anthropic on AWS Bedrock as external AI services, and the GDPR declaration places the web application on AWS in Virginia. The privacy policy lists permitted categories of subprocessors rather than every provider, and no outside counsel disclosure pack is offered.
Court Disclosure Support
If a judge’s standing order requires an AI disclosure, can the product produce one?
No located public material addresses court disclosure or verification certification.
Patent records and claim charts export in spreadsheet, document, XML and CSV forms for review, which supports invalidity and litigation work. No court disclosure export records which parts of a chart the AI produced or the prompts behind it.