D

Docket Alarm

Talk to the Docket answers plain English questions about a case, and AI Summaries condense a filing on opening. Extraction turns docket entries and filings into the normalized party, counsel and entry structure that the API returns and the analytics run on. Take all of that away and what remains is the business Docket Alarm has sold for a decade and still leads with: full text search across a billion filings, alerts, docket tracking, direct PACER retrieval and bulk download.

The published plans are priced in units of the non AI product, at 4 dollars a document, 3 a docket update, 2 a day for tracking and 1 a search. The analytics products are sold separately as the Federal Analytics, Easy Analytics and Analytics Workbench add ons.

The AI Assistant page says the system draws only on the docket being viewed and reads the actual record from the complaint through the latest filing, including the documents behind the entries. Every statement links to its supporting entry or document, so a reader can open the source in one click. The page includes a worked example: an answer on a named 2026 Northern District of California declaratory judgment case, with a Sources list identifying the complaint and its filing date.

No accuracy rate, test set, evaluation or error analysis is published. The same page states as an absolute that output carries no hallucinated facts or fabricated citations, while the vendor's own agreement says machine generated output is probabilistic and should be evaluated for accuracy, including by expert human review. The system retrieves case records rather than stating propositions of law.

Each statement in an answer links to the entry or document that supports it, and the vendor says the point of that link is to let the reader decide whether the answer is correct. Where a fact is not in the record, the system is said to say so rather than fill the gap. The agreement puts supervision in writing, telling the customer to evaluate machine generated output for accuracy, including through expert human review.

Nothing on the home page, AI Assistant page, API page or in the terms describes a confidence threshold or exception behavior, and there is no route to report or correct a wrong answer. The case briefing function, which the vendor says reads hundreds of docket entries and traces the procedural history unattended, has no described checkpoint between the request and the finished brief.

No production evidence is published. The home page, the use cases page covering all five buyer segments, the about page, the AI Assistant page, the API page, the coverage page and the pricing page carry no named customer, attributed testimonial, case study, customer count, logo strip or figure for what changed at any organization. In their place are unattributed claims: that thousands of legal professionals rely on the platform, that the vendor has relationships with many large firms, and that it has direct partnerships with law schools, none of them named.

The five example workflows on the use cases page are explicitly hypothetical. The one time saving published, that a case brief replaces two to four hours of paralegal work, is a product claim with no customer or method attached. The platform has been in market since well before its 2018 acquisition.

The privacy policy recognizes that litigation work involves viewing material about adverse parties, and states that personal search history is kept strictly confidential to the user's organization and not sold to third party marketers. The terms of service, which govern, run the other way. They reserve the right to monitor and record usage, including content interactions, search history and IP addresses, and to track activity to learn more about the customer and its users.

They grant the vendor an irrevocable, perpetual and sublicensable license over user content to improve its products. They disclose that a limited number of staff and third party foundational AI models may technically access user content. And the confidentiality clause excludes publicly available information, which covers nearly all of the court records in the product. Nothing reconciles the two documents. No training position, retention period, access model within a firm or treatment of privilege or work product is published. The terms, not the privacy policy, set the position.

A clause in the terms says no legal counsel is provided: content is for practical and informative purposes, is not professional advice, creates no attorney client relationship, and users should seek qualified counsel before acting on it. A separate clause calls AI output machine generated probabilistic prediction to be evaluated for accuracy, including by expert human review, which addresses competence and supervision.

The audience is defined. The services are for professional use, and consumers and nonprofessionals are expressly barred from them, so there is no consumer surface to disclose on. No bar or ethics guidance is named, including ABA Formal Opinion 512. No limit by practice jurisdiction is stated. The only geographic limit is a license restriction on using the service outside the country in the subscription agreement, which is a commercial term rather than a statement about where the answers can be relied on. None of this appears on the home page, AI Assistant page, use cases page, about page or footer.

The Clio trust center, public at trust.clio.com, has an AI section with three named topics (AI Risk Management, Third-Party AI Diligence and Employee AI Usage) and a document titled AI Overview. Docket Alarm's own site has nothing. None of the home page, AI Assistant page, use cases page, about page, coverage page, API page, privacy policy or terms has a responsible AI statement or an accountable owner. None describes testing before release or any evaluation of fairness or uneven output, and no Docket Alarm page links to the trust center.

The vendor's corporate use case invites customers to search every court for a person's full litigation record before hiring an executive or onboarding a vendor, and to judge whether that person is litigious and how their cases tend to end. Nothing published addresses screening named individuals on litigation history.

The privacy policy names Google App Engine as the host for all data except payments, and uses a PCI certified payment processor with card details encrypted with AES-256 and no employee access. It commits to seek consent before using information for a new purpose and offers prior versions on request. The removal route is detailed. Suppression requests are free, with a stated five business day decision window. Criteria are published both ways: cases mentioning minors are suppressed automatically, while cases of strong public interest and requests from legal professionals who want to be disassociated are refused.

The vendor also notes that search engine snippets may persist for weeks because it does not control them. The policy and the terms state no retention period, no deletion route for customer account data and no access control model. They give no encryption statement beyond payment details, no subprocessor list, no breach notification practice and no date. The parent trust center publishes incident response, breach notification and backup material, but it is not linked from this site and its coverage of Docket Alarm is not stated.

The terms cap total liability, whatever the cause, at one hundred dollars, and exclude indirect, special, incidental and consequential damages, including lost profits, revenue, data and opportunities. Named carve outs sit outside the limit: fraud, willful misconduct, violations of law, infringement or misappropriation of intellectual property, and breach of confidentiality. Warranties are expressly disclaimed on an as is and as available basis, including merchantability, fitness for purpose and noninfringement, with specific disclaimers that results will be accurate, reliable or free of errors.

A separate clause declines to warrant AI output. The contracting entity, governing law and forum are named. Nothing runs the other way. The only indemnity runs from the customer to the vendor, no warranty attaches to output, and the agreement has no insurance position, though the parent trust center lists a liability insurance document.

The developer API page names its endpoints with methods and describes what each returns. Search and Search PACER query by party, attorney, court, date range, case type or full text. Get Docket returns the full docket sheet with every entry, parties, counsel, related cases and status. Get Document returns the PDF behind an entry. Track a Case is a POST that subscribes to a case or saved search, and Ask Docket exposes the AI assistant's grounded answers.

A sample JSON response is published, and a full API reference sits at a public documentation path. The terms name three native integrations: Microsoft OneDrive, Dropbox and PACER ECF accounts. No practice system is named. The API page invites a firm to feed dockets into its matter management, DMS or BI stack but offers no ready connector, and nothing is published for iManage, NetDocuments, Word, Outlook, ebilling or filing. Data flows one way, out of Docket Alarm, with no sync back described.

The privacy policy states that all data other than payment information is stored on servers run by Google through Google App Engine, and points to Google's own privacy material. The home page, API page, coverage page, terms and privacy policy state no region, residency option or tenancy model, and nothing separates where data is processed from where it is stored. The ownership layers disagree. Docket Alarm's privacy policy names Google App Engine, while the parent trust center's infrastructure section names Amazon Web Services and publishes separate status pages for the United States, Canada, the European Union and Australia. Neither site says which applies to this product.

The Clio trust center is a SafeBase portal open without a sales conversation. It names GDPR, HIPAA, SOC 1 Type 2, SOC 2 Type 2 and TX-RAMP, and states that the 2025 SOC 2 Type II report covers 1 June 2024 to 31 May 2025. It offers a penetration test summary, both SOC reports, a liability insurance document and an AI overview through a self serve access request. The portal's overview names the Vincent and vLex products, not Docket Alarm, and no page on docketalarm.com links to it; the footer carries only privacy, terms and case removal.

The home page, AI Assistant page, use cases page, about page, API page, coverage page and both legal documents carry no security page, named certification, auditor, scope statement, coverage period or trust portal of Docket Alarm's own. Nothing published places the product inside an audited boundary.

The terms state that a limited number of staff and third party foundational AI models used in the vendor's intelligent technology services may technically access user content. So customer material reaches an external model that is not named. Nothing on the AI Assistant page, the API page or any other Docket Alarm page names a model, provider, version or processing location, and no commitment to notify customers of a model change is published.

The parent company does publish a subprocessor list naming Google Cloud, OpenAI, Anthropic and Amazon Web Services. But that portal says it covers the Vincent and vLex products, does not name Docket Alarm and is not linked from this site, so nothing ties those providers to Talk to the Docket or AI Summaries.

Two self serve plans are published with figures. Flat-Fee costs 99 dollars a month and covers unlimited document access, case alerts, docket tracking and search, plus motion filtering. Pay-As-You-Go costs 39.99 dollars a month, with every unit priced: 4 dollars a document, 3 a docket update, 3 a day for new case alerts, 2 a day for docket tracking and 1 a search in federal courts. Three add ons are named: Federal Analytics, Easy Analytics and the Analytics Workbench.

Pass through costs are explained plainly. Government fees are passed on at the government's own rate with no markup, and the PACER schedule is reproduced at 10 cents a page with a 30 page or 3 dollar cap and no cap on transcripts. Nothing is charged for a document already held, and state courts that charge fees, mostly in California, are always offered as a choice to accept or decline at the point of access. A free tier gives a stated weekly allowance of searches and document views for material already in the system, and a three day trial with no commitment is offered. API pricing goes through sales and has no published figure or unit.

The use cases page treats five buyer segments separately, with the question each brings and the features that answer it: litigators (docket tracking, opposing counsel and judge analytics), IP counsel (USPTO, PTAB, TTAB and ITC Section 337 work), law students (real filings and law school programs), in house counsel (portfolio litigation risk and agency enforcement) and corporate teams (counterparty and vendor screening).

The API page adds four more, and the signup form lists nine organization types, including government, school and bar association. A coverage page lists courts one by one: all 94 federal district courts, all 13 circuits, every bankruptcy court and the Supreme Court, stated as complete PACER coverage. Agency tables by practice area name the USPTO, PTAB, ITC, FDA Orange Book, TTAB, US Tax Court, FERC, FTC and NLRB, each with a coverage status.

State courts are listed at more than 43, and international coverage by country. The state map legend includes a not yet covered category, and a route to request a court is published.