Alt Legal
Alt Legal is trademark docketing software for IP practices. It connects to more than 180 intellectual property offices, derives matter data automatically from official filings, calculates due dates, and keeps a global portfolio synchronised without manual entry. Alongside docketing it sells trademark protection tools, including a Section 2(d) watch that flags conflicting applications before publication and an automated global application match service. Two search and drafting aids sit inside the platform: the Goods and Services Composer, which brings the full USPTO ID Manual together with two million live registered marks and shows filing fees by class as a description is built, and the AI Filtering Assistant, a natural-language search tool that turns a typed request such as a jurisdiction and date range into the filters behind a report, and which an administrator can switch off. The company also offers Alt Legal Assist, a service in which its own IP paralegals manage a customer's docket under a separate agreement. Subscriptions are priced on the number of docketed matters, start at $60 per month, and include unlimited users, data migration, training and support at no additional cost. Alt Legal, Inc. is a Delaware corporation with offices in New York and London, and it acquired the WebTMS global IP portfolio management product in April 2026.
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.
One shipped feature uses machine learning: the AI Filtering Assistant, a natural-language query tool that populates report filters from a typed request such as patent filings in the EU from the last 90 days. Everything else the vendor labels intelligent turns out on inspection not to be. The Goods and Services Composer is described as an intelligent drafting tool but its published mechanism is search and reuse over the USPTO ID Manual and two million registered marks, with snippet libraries, deduplication and fee calculation, and no model is claimed for it. Automated docketing, the core of the product, is rules-driven date calculation against official IP office data. Alt Legal states plainly on the AI page that it does not train or host its own AI models, and the docketing platform's value stands entirely without the assistant.
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.
Accuracy is claimed and never measured. The Composer page promises improved ID Manual compliance and accuracy, and the AI page says operational data including queries and AI responses is logged to monitor quality and reliability, but no accuracy figure, test set, error rate or failure mode is published on any surface read on 31 August 2026. Two limbs of this axis do not bite: the product retrieves no primary law and generates no citations to authority, so there is nothing to ground or to check for subsequent history. The failure mode that does apply is a misconstrued filter returning an incomplete docket report, and against that the design helps more than the disclosure does, because the assistant populates visible filters in the report the user is already in rather than returning an opaque answer.
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.
The AI feature is user-invoked and non-agentic: it runs when someone types a query and it changes filters rather than acting on a matter. There is one real published control, and it is an administrative one, with the AI page stating that the assistant can be enabled or disabled according to a customer's IT and security preferences. What is missing is any statement of what a user should check before relying on the output, or any described review point. The nearest thing is a term in the customer agreement placing responsibility for reviewing the accuracy of all output on the customer and its authorised users, which allocates the duty without describing a mechanism for discharging it.
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.
The customer evidence is real, attributed and detailed, and it contains no measurement of change. The Nolte Lackenbach Siegel study names the firm, its managing partner Renee Duff and its trademark paralegal Kim Hassiak, and describes a portfolio of nearly 7,000 active USPTO matters and nearly 6,000 international matters at a firm it identifies as a top ten US trademark filer, along with the merger and legacy-system failure that drove the migration and the specific features the team uses. Further named practitioners appear across the site with firms and titles, including Rosie Burbidge at gunnercooke and Hillary Schroeder at Name Game IP, against a portfolio claim of 2,000 law firms and legal departments. What no located study provides is a before and after figure: the numbers describe portfolio scale rather than what changed, no study carries a publication date, and none of the evidence is attributed to the AI feature.
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.
The disclosure for the AI feature is the most specific located in this pull. The AI page states that only the query, system instructions and structured filtering data are sent to the model, that customer records and portfolio data do not leave Alt Legal's controlled infrastructure, that the AI provider does not store prompts or responses, and that logged operational data is never used to train models. Behind that sits a contractual confidentiality regime: section 3.1 of the terms defines the customer's confidential information to include information pertaining to legal or business affairs, clients, and other non-public data. Three gaps keep it here. Privilege and work product are never addressed, and the separate Alt Legal Assist agreement states expressly that Alt Legal cannot opine on privilege. Nothing addresses segregation between matters or between users inside a firm. And the terms and privacy policy pages refuse automated access, so both were read only in the fragments recoverable through the search index.
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.
The advice line is drawn in the agreement rather than in a footer. Section 2.3 of the terms states that Alt Legal does not act as a law firm or a lawyer and does not act as a substitute for the advice of a lawyer, and that nothing about the availability, operation or use of the platform is intended to create an attorney-client relationship. The Alt Legal Assist agreement repeats the point for the paralegal service, adding that Alt Legal cannot opine on privilege or confidentiality. That is a real position, stated twice, in the documents that bind. It stops short of the rest: nothing engages with ABA Formal Opinion 512 or any state bar guidance, nothing addresses a supervising lawyer's competence and supervision duties over either the software or the outsourced paralegals, and no jurisdiction limit is stated for a product sold into 180-plus IP offices.
AI Governance and Bias Disclosure
Published governance over model behaviour: who owns it inside the vendor, what is tested before release, and what is disclosed about disparate output across matter types, parties, or populations.
Searched the home page, the AI Filtering Assistant page, the Goods and Services Composer page, the pricing page, the GDPR page, the Nolte case study and the blog index on 31 August 2026: no responsible AI statement, no governance framework, no named owner for AI decisions, no pre-release testing regime and nothing on bias or uneven output was located. The AI page's Security and AI Infrastructure section is substantial but it is about data handling and access control, which is a different subject and is credited on the stewardship and confidentiality rows rather than here. The vendor's own statement that it does not train or host models explains part of the absence, since it is not making model-level decisions, but it does not answer what is tested before a feature that reads across a firm's whole docket is released.
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.
Most of the ground is covered and the artifacts are not published. The GDPR page states TLS 1.2 or greater in transit, encryption at rest described as SHA-256 with 2048-bit keys, backups at least daily, annual employee privacy training, completed data mapping and privacy impact assessments, and documented incident response procedures. It sets out a deletion route conditioned on processing being complete and retention obligations satisfied, and it acknowledges Alt Legal's role as a data processor for GDPR customers. For the AI feature specifically, access to logged queries and responses is stated to be restricted to authorised engineering and security personnel. The gaps are the ones this band names: no subprocessor list is published, being available only by emailing the team, no retention period is stated for anything, and no breach-notification commitment was located. The GDPR page was last modified in May 2022 and the privacy policy could not be fetched because the path refuses automated access.
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.
The agreement allocates the loss and offers no recourse. Section 5 caps Alt Legal's maximum liability at the aggregate fees paid in the six-month period before the claim arose, which is at least a specific and readable measure, and excludes special, indirect, incidental, punitive and consequential damages along with lost profits and attorneys' fees. The platform is supplied AS IS with warranties disclaimed, and the same section states that the customer and its authorised users are solely responsible for reviewing the accuracy of all information they input and any output the platform generates. Authorised users also waive rights of setoff and counterclaim other than compulsory ones. No indemnity running to the customer, no warranty on output and no insurance position was located in the recovered text, though the terms page refuses automated access and was read only in fragments, so the absence of an indemnity is not asserted as a finding.
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.
Real connections exist and none is documented to a depth an implementer could use. The pricing page's feature list names TEAS integration for USPTO filing, calendar sync for deadlines, client and third-party docket access, intake forms and data for 180-plus IP offices, and the Nolte study describes a working link to Smartsheet that Alt Legal's support team helped build at no additional fee. What was not located on 31 August 2026 is an integrations directory, an API reference or any developer documentation, and no connection to a document management system such as iManage or NetDocuments was found. The 180-plus office connections are the product's data supply rather than integration into the systems a firm's other work lives in, and they are credited on coverage rather than here.
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.
Where processing happens is answered and how the software is tenanted is not. The AI page states that all AI processing takes place within a private US-based cloud environment and that customer records and portfolio data do not leave Alt Legal's controlled infrastructure, which is a more precise processing-location statement than most vendors in this pull publish. The privacy policy adds that personal data may be transferred to and stored outside the EEA, including in the United States, and the GDPR page identifies Alt Legal, Inc. as a Delaware corporation based in the United States. Against that, no tenancy model is described anywhere, no region options are offered, and no EU or UK hosting alternative is mentioned despite a London office and European customers. A buyer can learn the country and not the architecture.
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.
No independent security attestation was located. Checked the home page, the AI page, the pricing page, the GDPR page and the site footer on 31 August 2026, and ran a separate search for a SOC 2 or ISO 27001 report in this vendor's name, which returned only generic explanatory material about the two standards and nothing naming Alt Legal. There is no trust portal, no badge, no named auditor and no report request route. The GDPR page does describe security practices in specific terms, including transport security, encryption, backup cadence and incident response procedures, but a description of controls by the vendor is not an attestation by anyone else, and this axis asks who audited them.
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.
The vendor describes the arrangement clearly and will not say whose model it is. The AI page states that Alt Legal uses a secure enterprise AI platform to power the assistant, that it does not train or host its own models, that processing happens in a private US-based cloud environment, and that the AI provider does not store prompts or responses and does not use customer data for training, advertising or profiling. Every one of those facts is about the relationship rather than the counterparty: no model, no model family, no vendor and no version is identified anywhere on the surfaces read on 31 August 2026, and no commitment to notify customers if the provider changes was located. A firm answering a client's question about which model sees its docket queries cannot answer it from published material.
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.
This sits at the top of the band and stops short of the range. Four tiers are published with figures against a stated unit of total docketed matters: up to 50 matters at $60 per month, up to 100 at $100, up to 200 at $195 and up to 400 at $295, each with the optional Trademark Protection add-on priced alongside at $30, $40, $60 and $95. A docketed matter is defined in the page's own FAQ as a distinct trademark, patent or opposition filing. What implementation adds is answered directly and unusually: no setup fee, free bulk data migration completed within about a week, and unlimited support and training, with unlimited users on every tier. Above 400 matters the page invites contact, so the larger portfolios the vendor's own case studies feature, including one at nearly 13,000 live matters, are unpriced. One inconsistency worth noting: the page's meta description still advertises a $50 starting price against a table that reads $60.
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.
The product says what it is for with unusual candour. It is built for trademark practice and the vendor does not pretend otherwise: the Nolte study describes that firm keeping its patents on a separate patent-focused system while trademarks run on Alt Legal, and names the approach a best-in-class split docket, with the CEO publishing an article arguing the case. Supported matter types are itemised beyond trademarks, covering TTAB oppositions and cancellations, copyrights, disputes, licences, patents, state registrations and custom types, across 180-plus jurisdictions, and the buyers named run from solo practitioners through boutique firms to in-house departments, with a separate track for law schools and students. Boundaries are stated for practice areas but not for segments: no firm-size floor or ceiling is given, and government use is not addressed.
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?
A public policy or trust page states no training on customer content, with no matching term located in the published agreement.
The AI Filtering Assistant page makes the commitment twice, once for the external provider in the quoted line and once for Alt Legal itself, stating that logged operational data is never used to train AI models. It also limits what reaches the model at all, saying only the query, system instructions and structured filtering data are sent while customer records and portfolio data stay inside Alt Legal's infrastructure. No matching prohibition was located in the customer agreement, which refuses automated access and was read only in fragments, so this is recorded as a policy commitment on a product page rather than a contractual one.
Prompt and Output Retention
How long does the product keep what a lawyer typed, and can that be set to zero?
Retention is acknowledged in public materials with no stated period.
Retention is split between the two parties and only one half is answered. The provider stores nothing, per the quoted line. Alt Legal itself does retain material, stating that it logs limited operational data such as queries, AI responses and performance metrics to monitor quality, reliability and security, with access restricted to authorised engineering and security personnel. No period is stated for those logs, no customer control over the window is described, and no zero-retention option is offered.
Ethical Walls and Matter Segregation
Does retrieval respect the firm’s ethical walls, or can the model read across them?
No located public material addresses walls or matter level segregation.
No located material addresses ethical walls, matter-level segregation, or whether the AI Filtering Assistant respects per-user access restrictions when it queries across a firm's docket. The assistant is described as searching across all of a user's matters, and the only access control published is an administrative switch that enables or disables the feature for the organisation as a whole. Searched the home page, the AI Filtering Assistant page, the Composer page, the pricing feature list, the GDPR page and the Nolte case study on 31 August 2026; the terms and privacy policy pages refuse automated access and were read only in fragments.
Third Party Request and Subpoena Notice
If someone subpoenas the vendor for a firm’s data, does the firm hear about it first?
Terms commit to notice where lawfully permitted. No transparency report located.
The quoted commitment appears in the Alt Legal Assist terms of service, which govern the paralegal service rather than the docketing platform, and it runs further than most: on receiving a judicial or government order to disclose the other party's confidential information, the receiving party must give notice and a sufficient opportunity to contest the order, to the extent legally permissible. Whether the main platform agreement carries an equivalent clause could not be established, because its confidentiality section was recovered only in fragments from the search index after the terms page refused automated access. No transparency report was located.
Primary Law Corpus Provenance
Where does the law in this product come from, and does the vendor have the right to use it?
Sources are identified without stating the licence or rights basis.
The product does not retrieve case law or statutes, so the corpus behind it is registry data rather than primary law, and that corpus is named specifically: the full USPTO ID Manual and roughly two million live registered marks in the Goods and Services Composer, and filing data from more than 180 IP offices for docketing. Currency is claimed, with the Composer said to stay in sync with current USPTO-acceptable language across all 45 NICE classes. No licence, rights basis or public-domain statement is given for any of it, and the arrangements for non-US offices are not described.
Good Law Verification
Does the product tell you when the authority it just cited has been overruled?
No located public material addresses whether authority is checked for subsequent history.
No citator applies, because the product does not return legal authority whose subsequent history could be checked. The nearest analogue is status tracking rather than treatment: docketing pulls live status changes from IP office records, so a registration's current standing is maintained automatically, but that is registry currency rather than a good-law signal. Searched the home page, the Composer page, the AI page and the pricing feature list on 31 August 2026.
Refusal and Uncertainty Behaviour
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.
Nothing located describes what the AI Filtering Assistant does with a query it cannot parse or cannot satisfy. No abstention path, no confidence indicator and no error behaviour is published. The design does give the user something to check, since the assistant populates visible filters in the report rather than returning a bare answer, but the vendor does not describe this as an uncertainty mechanism. Searched the AI Filtering Assistant page, the home page and the pricing feature list on 31 August 2026.
Fabricated Citation Record
Does a public court record exist involving output from this product?
No court order, opinion or disciplinary record naming this product has been located as of the date shown. This is a statement about the public record, not a finding about the product.
Searched the AI Hallucination Cases database maintained by Damien Charlotin at HEC Paris, together with 2026 sanctions trackers and trade coverage, on 31 August 2026 on both the product and company name, which are the same. No court order, opinion or disciplinary record naming Alt Legal was located. This is a statement about the public record rather than about the product, and the failure mode this signal tracks fits it poorly: the AI feature builds report filters over a firm's own docket and generates no citations to authority.
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.
No engagement with ABA Formal Opinion 512 or any state bar guidance on generative AI was located. The vendor publishes a substantial professional education programme, including accredited CLE, a webinar series on trademark law topics and a paralegal course and community, but none of the located material addresses the ethics obligations attaching to a lawyer's use of AI. Searched the home page, the AI page, the events and CLE pages and the blog index on 31 August 2026.
Billing and Fee Posture
Does the vendor address what happens to the bill when the work takes an hour instead of six?
No located public material addresses billing, fee or disclosure treatment.
Nothing located addresses how AI-assisted work should be billed or disclosed to a client. The vendor does discuss fees elsewhere, describing custom tags used to distinguish which clients are billed for which tasks and noting that clients are not billed for checking the docket themselves, but that is client billing administration rather than the fee treatment of AI-assisted work. No record of what the assistant did on a given matter is described. Searched the AI page, the pricing page, the Nolte case study and the blog index on 31 August 2026.
Outside Counsel Guideline Readiness
Can a firm get this vendor through a client’s AI clause without a bespoke negotiation?
The material exists behind a sales conversation or an executed agreement.
The GDPR page states that a customer wanting details of Alt Legal's sub-processors or a copy of its data protection agreement can obtain them by emailing the team, calling, or using live chat, and says the vendor regularly reviews its vendor agreements and data protection agreements. Both artifacts therefore exist and neither is published: they sit behind a human request rather than a self-serve portal or download. A firm could not assemble a client-facing AI disclosure from public material either, since the model provider behind the AI feature is not named anywhere.
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.
Nothing located addresses producing an AI-use disclosure or a verification certification. No per-document or per-matter export covering which queries were run, what the assistant returned or who reviewed it is described. Searched the AI Filtering Assistant page, the pricing feature list covering reporting, and the home page on 31 August 2026. The obligation is a poor fit for this product, which builds docket reports rather than court filings, but the row is recorded rather than skipped.