Product and intended use
Define the configuration, people, information and tasks being assessed.
What this profile covers and how AI Docketing is sold
Public claim only
AI Docketing is listed as a product in its own right in the products menu on every Anaqua page read, alongside AQX, RightHub, PATTSY WAVE, Patricia, Unified Patents, Brand Protection and others, and it has its own product page at /products/ai-docketing/, which the site's sitemap.xml also lists. The vendor states it is delivered on two platforms: 'Anaqua's AI capabilities are delivered across our platforms - AQX provides enterprise-scale IP management and RightHub delivers connected, AI-native workflows', and it names AI docketing in the feature list of each. Nothing read states whether AI Docketing is included in an AQX or RightHub subscription, sold as an add-on module, or priced by document, seat or volume.
Ask: Is AI Docketing an included capability of AQX and RightHub or a separately licensed module, and can it be bought without either platform?
Published functions
Public claim only
The product page states the product will 'Turn a flood of PTO correspondence, counsel emails, and multi-jurisdictional filings into structured, docketed records with AI — without re-keying a single field.' Named functions are: multi-source intake, bringing documents 'from PTO integrations (USPTO, EPO, EUIPO), forwarded emails, and uploads into a single queue — including AI processing for JPO and CIPO documents'; automated document and email processing, where 'each item is classified, data and deadlines extracted, and updates proposed to the correct record'; intelligent case matching 'using references, application numbers, and jurisdictional data', with unmatched items surfaced 'for manual assignment'; deadline and event automation, to 'Capture deadlines from document content, create docket events, manage completed tasks, and trigger downstream workflows'; prior art and IDS propagation, to 'Retrieve prior art PDFs in bulk from USPTO 892 forms, propagate citations across patent families'; and 'Full auditability & traceability' through 'a complete audit trail'.
Ask: For each function, what is written to the docket automatically and what is only proposed, and is the audit trail exportable?
Release stage and availability
Not confirmed
No release date, version, changelog or availability statement is published for AI Docketing on the pages read, and no region or platform-version prerequisite is stated. The product page's related resources date the underlying document auto-processing work: a video titled 'AQX 11: Introducing Document-Auto Processing' dated 31 July 2024 and a data sheet 'AQX Document Auto-Processing' dated 16 June 2025 are listed, and a case study on AQX automation is dated 6 May 2024. Whether AI Docketing as now marketed is the same capability renamed, or a later product, is not established by anything read.
Ask: When was AI Docketing released as a named product, is it the same capability previously marketed as AQX Document Auto-Processing, and is it generally available on both AQX and RightHub today?
How it is delivered and where it runs
Not confirmed
Nothing on the AI Docketing page addresses hosting, region or deployment; it is delivered inside AQX or RightHub. The host platform page for AQX states 'Built on Microsoft Azure', 'Azure-based, single-tenant hosting for maximum isolation and resilience' and, separately, 'U.S.-only infrastructure and staffing for sensitive and export-controlled data'. No Azure region is named anywhere, no data-residency option is published, and nothing read states whether AI Docketing runs in the same tenancy as the platform or whether document processing and model inference happen elsewhere. No region is inferred from the U.S.-only option or from the Azure reference.
Ask: In which Azure regions does AI Docketing store documents and run model inference for an Australian customer, is an Australian region available, and does document processing leave the platform tenancy?
Jurisdictional coverage, including Australia
Public claim only
The published coverage is office-by-office and Australia is not in it. The FAQ states 'You can process 1,000+ PTO document types across USPTO, EPO, EUIPO, JPO, and CIPO — including office actions, confirmations, grants, search reports, and general correspondence', and the intake function names 'PTO integrations (USPTO, EPO, EUIPO)' with 'AI processing for JPO and CIPO documents' as a separate category. IP Australia is not named on any page read, nor is WIPO, nor New Zealand, and no Australian deadline or docketing rule set is mentioned. The vendor's published statistics cover the same two offices: '95% of USPTO forms covered' and '99% of EPO forms covered'.
Ask: Is IP Australia correspondence supported for classification, extraction and case matching, and if so at what coverage, or must Australian office actions be docketed manually?
Vendor content directing how AI describes the product
Not confirmed
Checked deliberately on 2026-09-27 for a page, file or block telling AI models, assistants or crawlers how to describe the product. The complete sitemap.xml for anaqua.com was read: its top-level pages are /ai-overview, /anaqua-blog, /client-experience/, /codes-of-conduct/, /contact-us/, /cookie-policy/, /ip-platform, /leadership/, /legal-terms/, /our-culture/, /our-offices/, /our-story/, /privacy-policy/, /resources, /services/ and a few thank-you and style-guide pages, with no /llms.txt, /ai-info or /for-ai entry. The page named /ai-overview was read on the chance that the name indicated content addressed to models; it is a press-release and resources landing page addressed to readers. Neither the AI Docketing page nor the Anaqua AI page contains any block addressed to AI models or crawlers.
Ask: Do you publish any page, file or metadata addressed to AI assistants, answer engines or crawlers setting out how Anaqua products should be described?
Data, controls and requirements
Compare the actual configuration with firm, client and jurisdiction requirements, supported by evidence.
Client authority and confidential client correspondence
Not confirmed
The intake the vendor describes is client and counsel correspondence: 'Forward emails from outside counsel, foreign associates, clients, or internal teams into AI Docketing. The system reads the attachments'. For a law firm that means privileged client correspondence and unpublished filings are read by the system, and for a corporate IP team it means correspondence from external counsel. The vendor's guiding principles state 'Protect sensitive IP data with access controls and safeguards across data handling and outputs', which is a statement of intent. Nothing read states whether email content and attachments are sent to a model outside the customer's tenancy, or what notice a firm would need to give clients whose correspondence is processed.
Ask: Are forwarded emails and their attachments sent to any model or service outside our tenancy, and what do you expect us to tell clients whose privileged correspondence is processed?
Foundation-model providers
Not confirmed
No model, model family, version, platform or supplier is named on the AI Docketing page, the Anaqua AI page, the legal terms or the privacy policy. The AI Docketing page attributes the work to 'AI' throughout without qualification, and the FAQ answers 'How accurate is the case matching?' with 'AI Docketing uses high-accuracy matching based on application numbers, references, and jurisdictional data' without naming a technique or a model. The only related published fact is that AQX is 'Built on Microsoft Azure', which does not establish which model service performs inference or who supplies it.
Ask: Name every model and model provider used for classification, extraction, case matching and deadline capture, the legal entity supplying each, and the region of each inference endpoint.
Subprocessors
Not confirmed
No subprocessor list, register or trust centre is published, and this is now established against the whole site rather than a sample: the complete sitemap.xml for anaqua.com was read and its only legal documents are /legal-terms/, /privacy-policy/, /cookie-policy/ and /codes-of-conduct/. The global privacy policy refers to 'Anaqua, Inc. and its global affiliates' and to third parties who assist with developing, promoting and providing software and services, states that personal data 'may not be used by the third parties for any other purpose', and refers to 'Anaqua and/or its hosting providers' without naming one. Microsoft Azure is named on the AQX page as the platform it is built on.
Ask: Provide a dated subprocessor register for AI Docketing naming every entity including any model provider, its country of processing, what it processes, and our notice and objection rights.
Training on customer content
Not confirmed
Nothing read states whether customer documents, forwarded emails, extracted data, docket records or corrections are used to train, tune or evaluate any model. The Anaqua AI guiding principles say only 'Protect sensitive IP data with access controls and safeguards across data handling and outputs'. The AcclaimIP Terms & Conditions, which do not govern this product, point the other way for that one: they assign to Anaqua all right, title and interest in the software and its updates and modifications 'whether suggested, created, made, or provided by Anaqua, Customer, or any other person', and require the customer to assign any such rights it acquires.
Ask: Does any signed document prohibit training, fine-tuning, evaluation or product improvement on our PTO documents, forwarded emails and docket records, by Anaqua and by every model provider?
Use of corrections and validation decisions
Not confirmed
The product's core interaction produces correction data: items are proposed and a person validates them, and 'Items that can't be confidently matched are surfaced in an unmatched queue with suggested partial matches — so you always have visibility and can assign manually.' Every manual assignment and every rejected proposal is therefore a labelled correction. Nothing published says whether those corrections are retained beyond the customer's own tenancy, whether they are used to improve matching for other customers, or whether the customer owns them.
Ask: When our team rejects a proposed docket record or manually assigns an unmatched item, is that correction used to improve the model for other customers, and who owns it?
Retention
Public claim only
No retention period is published for documents ingested by AI Docketing, for forwarded email content, for extracted data or for the audit trail. The global privacy policy (updated 18 October 2023) states 'Anaqua retains PD only as long as necessary for Anaqua to comply with business, tax, and legal requirements', that for customers the period 'is likely to be the entire time you are a Customer', that after the relationship ends personal data is kept 'for a reasonable period', and that an anonymised form may be kept 'for statistical purposes without time limits'. That policy addresses personal data held by Anaqua, not a customer's docket content.
Ask: How long do you retain ingested documents, forwarded email content, extracted field data and the AI Docketing audit trail, and can we set that period?
Deletion on request and on exit
Public claim only
The privacy policy provides individual rights including 'The right to delete personal information that we have collected from you, subject to certain exceptions' and a right of portability, exercisable by email to a published privacy address with identity verification, and permits an authorised agent with a power of attorney or written authorisation. Those are individual rights over personal information. No deletion commitment, timetable or certificate is published for a customer's ingested documents, extracted data or audit trail on termination, and no such clause appears in the only published software terms, which govern a different product.
Ask: On termination, what is deleted from AI Docketing and its processing pipeline, on what timetable, and will you certify deletion including of any model artefacts derived from our documents?
Anaqua staff and contractor access
Not confirmed
Nothing published states which Anaqua staff or contractors can read a customer's ingested documents, forwarded emails or docket records, from which countries, or under what approval, and no customer-visible access log is described. The audit trail the product page describes is a record of processing actions for compliance and quality teams, not a record of vendor access. The offices page shows staff locations in the United States, France, Germany, Italy, Norway, the United Kingdom, Australia, India and Japan; the AQX page states a 'U.S.-only infrastructure and staffing' option 'for sensitive and export-controlled data', which implies staffing is otherwise not restricted by country but does not say where it is.
Ask: List every country from which your staff or contractors can access our ingested documents and docket records, and is that access logged and available to us?
Authentication and permissions
Not confirmed
The product page describes configurable oversight levels and an audit trail but no authentication or permission model. The FAQ states 'You configure your preferred validation level — from fully automated processing to one-step or two-step (four-eyes) review', which is a workflow control rather than an access control. Nothing read states whether single sign-on, enforced multi-factor authentication or SCIM provisioning is available, who may change the validation level, or whether that change is recorded in the audit trail.
Ask: Which single sign-on and multi-factor standards are supported, who can change the validation level from two-step review to fully automated, and is that change recorded in the audit trail?
Encryption in transit and at rest
Public claim only
Nothing on the AI Docketing page or the Anaqua AI page addresses encryption. The host platform page for AQX states 'Encryption at rest and in transit protects sensitive IP at every stage', with no algorithm, protocol version, key-management arrangement or customer-managed-key option named, and it is a marketing page with no contractual backing found. The privacy policy is weaker and refers only to 'security techniques designed to protect your information from unauthorized access, including firewalls and access control procedures'. Whether the AQX statement extends to the AI Docketing processing pipeline is not stated.
Ask: Specify the encryption in transit and at rest applied to documents while they are being processed by AI Docketing, including any queue or model endpoint, who holds the keys, and whether customer-managed keys are available.
Independent security assurance
Public claim only
No certification, audit report, penetration-test summary or trust centre is published for AI Docketing, and the complete sitemap for anaqua.com contains no security or trust page at all. The AQX product page states 'Certified security controls validated by ISO 27001, SOC 2 Type 2, and ISO 9001' without naming the certified entity, certificate number, auditor, scope or date, and the privacy policy states something narrower and differently scoped: 'Anaqua and/or its hosting providers are at least ISO/IEC 27001and/or SSAE-16/SOC 2, Type 2 certified'. So the certified entity is not established, and no document states whether the AI Docketing pipeline is in scope of any certificate. No ISO/IEC 42001 or AI-specific assurance is claimed anywhere. GDPR: the vendor's own page states "ction, use, and retention of personal information including the General Data Protection Regulation (GDPR), Switzerland, the UK, the California Consumer Privacy Act (CCPA), and other ap" (https://www.anaqua.com/privacy-policy/, checked 2026-10-10). ISO 27001: the vendor's own page states "nd are committed to performing the examination in future years." In November 2017, Anaqua achieved ISO 27001 certification, the leading international standard for measuring information sec" (https://www.anaqua.com/resources/anaqua-successfully-completes-soc-2-examination, checked 2026-10-10). SOC 2: the vendor's own page states "xamination, formally known as a Report on Controls at a Service Organization Relevant to Security. SOC 2 examinations evaluate controls at a service organization relevant to categories" (https://www.anaqua.com/resources/anaqua-successfully-completes-soc-2-examination, checked 2026-10-10).
Ask: Provide the current ISO 27001 certificate and SOC 2 Type II report for the entity that operates AI Docketing, with the scope statement and auditor, and confirm in writing whether the AI processing pipeline is in scope.
Contract, DPA, liability and forum
Not confirmed
No contract for AI Docketing is published. The legal-terms page carries two documents with different governing laws: website Terms of Use 'governed by and construed in accordance with the laws of the State of Delaware' with exclusive jurisdiction in Delaware courts, and the AcclaimIP Terms & Conditions for a different product governed by 'the laws of the Commonwealth of Massachusetts' with exclusive jurisdiction there, an express waiver of jury trial, and, per the sibling review of the same document, liability limited to fees paid in the preceding twelve months. Neither is stated to cover AI Docketing. No data processing agreement is published; the privacy policy states only that 'For the purposes of data protection laws, Anaqua's Customers are the "data controller" and Anaqua is the "data processor"'.
Ask: Provide the master agreement and data processing agreement that would govern AI Docketing for an Australian customer, naming the contracting entity, its ACN or ABN, the governing law, the forum and the liability cap.
Cross-border transfers and APP 8
Not confirmed
The privacy policy names no country of processing and addresses transfers only from the European Union: 'Anaqua complies with the EU Standard Contractual Clauses for all onward transfers of PD from the EU', and it states Anaqua may be required to disclose personal data 'in response to lawful requests by public authorities, including to meet national security or law enforcement requirements'. It lists the GDPR, Switzerland, the United Kingdom and the California Consumer Privacy Act as applicable laws and does not mention the Australian Privacy Act or Australian Privacy Principle 8. The only region named anywhere is the AQX page's 'U.S.-only infrastructure and staffing' option for sensitive and export-controlled data.
Ask: For an Australian customer, list every country in which our documents are stored, processed for inference, supported, logged and backed up, and state which APP 8 mechanism you rely on.
Incident notification and history
Not confirmed
No incident notification commitment, period, security contact, disclosure channel or history is published on the pages read, and the complete sitemap contains no security page on which one might sit. The privacy policy says only that Anaqua describes security measures 'to protect against the loss, misuse, and alteration of information' and 'cannot guarantee that an unauthorized disclosure will not occur'. The Australian Notifiable Data Breaches scheme is not mentioned, and no clause covers an incident at a model provider or hosting provider.
Ask: What is your contractual notification period from discovery of an incident affecting AI Docketing or its model provider, who is notified, and have there been any material incidents in the last 24 months?
Change notice for the model and the workflow
Not confirmed
No change-notice, release-note or deprecation commitment is published for AI Docketing, and no changelog was found. Nothing obliges Anaqua to notify a customer before changing the model behind classification, extraction or case matching, before changing which document types are supported, or before changing what a validation level does. The website Terms of Use address changes to the website terms only. The product page's supporting statistics, such as '95% of USPTO forms covered', carry no version or effective date against which a change could be measured.
Ask: Will you give advance written notice before changing the model or the supported document-type list, and where are AI Docketing release notes published?
Export and exit
Not confirmed
No export or exit provision for AI Docketing, AQX or RightHub is published, in any format, at any cost, with any post-termination availability period. The only published software terms point the other way for the product they govern: the AcclaimIP Terms & Conditions state 'Access of the Software in an automated manner is not permitted. The Data from the Software may not be directly or indirectly copied, disassembled, resold or redistributed in any manner', and reserve a right to ban users for automated access. Whether any equivalent restriction applies to bulk extraction of a customer's own docket records and ingested documents is addressed nowhere.
Ask: In what formats can we bulk-export our docket records, ingested documents, extracted field data and audit trail on exit, at what cost, and does any clause restrict automated extraction of our own data?
Connectors and data reach
Public claim only
The published reach is: inbound patent office integrations for USPTO, EPO and EUIPO; AI processing of JPO and CIPO documents; inbound forwarded email from 'outside counsel, foreign associates, clients, or internal teams'; manual uploads; bulk retrieval of prior art PDFs 'from USPTO 892 forms'; and outbound creation of docket events and triggering of 'downstream workflows' in AQX or RightHub. Anaqua lists a separate 'Application Connectors' product at /products/integrations/ which was not read, so the full connector set is not established.
Ask: Can AI Docketing be restricted to named intake sources and mailboxes, and does it reach any system outside AQX or RightHub?
Retrieval beyond the customer's own data
Public claim only
The product does reach outside the customer's own data, to public patent office sources: it takes documents 'from PTO integrations (USPTO, EPO, EUIPO)' and will 'Retrieve prior art PDFs in bulk from USPTO 892 forms, propagate citations across patent families'. Those are named public registers rather than open web search. No page read states that any component performs general web search, and none states positively that it cannot.
Ask: Beyond USPTO, EPO, EUIPO, JPO and CIPO sources, does any component query the open web or a third-party data provider, and can that be disabled?
Autonomy and action control
Public claim only
The vendor publishes fully automatic operation as a configurable option. The FAQ states 'Yes. You configure your preferred validation level — from fully automated processing to one-step or two-step (four-eyes) review — so your team sets the right balance of speed and oversight', and a feature block states 'Configure how much oversight your team applies — from fully automated AI processing to two-step review'. The page headline is 'Let AI own the intake. You own the outcome' and the standfirst states 'Configure your level of oversight and validate every action before it's confirmed', which sits alongside the FAQ's fully automated option; the page does not reconcile the two. In the fully automated setting, docket records and deadline events are created and downstream workflows triggered without a person confirming each one.
Ask: In the fully automated setting, what exactly can be created or changed without human confirmation — docket events, deadline dates, task completion, IDS citations — and which role can enable it?
Human review of AI output
Public claim only
Human review is available at two strengths and is not mandatory. The published options are 'fully automated processing', 'one-step' review and 'two-step (four-eyes) review'. Unmatched items are always routed to a person: 'Items that can't be confidently matched are surfaced in an unmatched queue with suggested partial matches — so you always have visibility and can assign manually.' Nothing read states the confidence threshold that decides what is treated as confidently matched, whether that threshold is visible or adjustable, or whether the reviewer is recorded in the audit trail.
Ask: What confidence threshold routes an item to the unmatched queue, is it visible and adjustable, and does the audit trail record which person validated each docket record?
