Concorda vs Litmas AI: how they compare in 2026
Concorda and Litmas AI both draft motions and pleadings for litigators from a case's own record and U.S. case law, and both check citations before the lawyer sees them. Concorda sits in the top two bands on ten of fifteen axes and Litmas AI on nine of fifteen, level on seven with four going each way. Concorda's lead is contract. Its terms and data processing addendum bar training on customer content, state that privileged material is processed only as a confidential service provider, commit to deletion within 30 days from production and 90 from backups with certification, and raise the liability cap to three times the fees for breach of the no training or confidentiality commitments. Litmas AI publishes no terms of service at all, and its commitments against training and on workspace isolation sit on its trust center and FAQ pages. Litmas answers on method and price. It describes a dual layer validation that removes any citation it cannot verify, names three firms that use it, and prices by the case at $125 a month.
At a glance
All 15 axes, side by side
The same grid applied to every vendor in the index, graded from public sources. 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.
The models are the product. Concorda drafts complaints, motions, briefs and discovery responses from the matter record, searches and verifies case law, and builds patent contention charts; the matter workspace that holds the record supports that work. Without the models there is little left to sell. Verified 22 September 2026.
The models are the product. Litmas AI answers case-specific research questions, extracts causes of action and maps evidence to their elements, builds a relationship graph of the case, drafts discovery and deposition questions, and drafts motions and pleadings from the case file and validated case law. The case library that holds the documents supports that work; without the models there is little left to sell. Verified 22 September 2026.
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.
Verification is claimed; the method is not described or measured. The home page says research runs across case law, statutes and local rules from every U.S. court in Concorda's own case database, updated daily and always citable to sources, and that every citation in the drafter is verified. How citations are verified, and how accurate the output is, is not published, and the terms warn that outputs may be inaccurate, incomplete or misleading and must be checked before use. Verified 22 September 2026.
The validation method is described; accuracy is claimed, not measured. The FAQ says every case, statute and pincite is checked against verified U.S. court opinions through a dual-layer validation system before it reaches the user, that a citation which cannot be validated is removed rather than shown, and that every citation that remains is linked inline so the lawyer can open it and judge its applicability. Facts are cited to the record at the pincite level. The pricing page promises zero fabricated case law, but no error rate, test set or independent evaluation is published, and case law is updated at a cadence that varies by court and is given on request. Verified 22 September 2026.
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.
Drafting happens in an interactive editor under a written review duty, short of stated limits. The drafter works collaboratively, with the lawyer directing rewrites, and the terms require the customer to independently review and verify every output before relying on, filing, serving or disclosing it. No mode, threshold or limit on what the product may do on its own is described. Verified 22 September 2026.
Guided workflows with the lawyer in the loop, short of stated limits. Drafting runs step by step, with verified facts assigned to a motion and elements marked as proven by the user, and the FAQ says the lawyer keeps full editorial control and can review every section before submission, and that the platform supports the duty to supervise rather than substituting for it. Nothing states what the product may not be used for or what happens when its output is wrong, and output is marketed as file-ready. Verified 22 September 2026.
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.
Checked the home, About, Plans and blog pages on 22 September 2026. No named customer, testimonial, case study or measured outcome is published. Verified 22 September 2026.
Named users without measured results. The site quotes Kristina Ducos of The Ducos Law Firm, George G. Leynaud of the Law Offices of Leynaud & Leynaud, and Richard R. Kennedy of the Law Offices of Richard R. Kennedy. The FAQ's claim that users typically cut motion drafting time by 60 to 80 percent gives no basis or source. Verified 22 September 2026.
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.
Written into the agreement, short of matter-level walls. The Terms and Data Processing Addendum bar using customer content to train, retrain or fine-tune any model; state that privileged material and work product are processed only as the customer's confidential service provider and are not intended to be waived, with confidentiality lasting as long as the protection does; send only limited data to third-party model providers under zero- or no-retention configurations, and never enable a provider that cannot meet that without disclosure and the customer's consent; and delete content within 30 days from production and 90 days from backups, with written certification. Separation between matters or users inside a firm is not described. Verified 22 September 2026.
Commitments on public pages, with no agreement behind them. The Trust Center and FAQ say case data is encrypted, isolated to the firm's workspace, never shared with other users, stored on U.S. servers and never used to train AI models, and the pricing table lists attorney-client privilege as protected. No terms of service, customer agreement or data processing addendum is published, nothing says which model providers process case data, and deletion after cancellation follows an unstated retention period. Verified 22 September 2026.
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.
A plain position in the agreement, short of jurisdiction limits. The Terms say AI outputs are informational, not legal advice; that the customer keeps sole professional responsibility, must independently review and verify every output before relying on or filing it, and remains responsible for its rules of professional conduct; that Concorda does not practise law and has no attorney-client relationship; and that the service is for organisations, not consumers. The jurisdictions the product covers or excludes are not stated. Verified 22 September 2026.
A clear position in the FAQ, with no terms behind it. The FAQ says Litmas AI does not provide legal advice or replace professional judgment, that the filing is the attorney's whatever tool produced it, that verifying output before filing remains the lawyer's responsibility, and that the platform is designed to support the duty to supervise. The product is built for litigators, and coverage is U.S. federal and state law. That position is not repeated in any published agreement, and the same pages describe output as file-ready and court-ready. Verified 22 September 2026.
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.
Checked the home, About, Plans and blog pages, the Terms and Data Processing Addendum and the Privacy Policy on 22 September 2026. No AI governance framework, accountable owner, testing before release, or finding on how drafting or research performs across practice areas or courts was located. Verified 22 September 2026.
Checked the home, FAQ, how it works, pricing, about and Trust Center pages and the Privacy Policy on 22 September 2026. No AI governance framework, accountable owner, testing before release, or finding on how research or drafting performs across jurisdictions, practice areas or case types was located. Verified 22 September 2026.
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 committed in the agreement; the subprocessor list is on request. The Data Processing Addendum sets out a written security programme with role-based access, multi-factor authentication, encryption in transit and at rest, logging, monitoring and incident response; deletion within 30 days from production and 90 days from backups, with certification; notice of a confirmed security incident within 72 hours with set contents and help with notifying clients, courts and regulators; and 30 days' notice before a new subprocessor, with a right to cancel. The subprocessor list, including AI model providers and processing countries, is available on request rather than published. Verified 22 September 2026.
Security claims are published; the policies behind them are not. The Trust Center says data is encrypted with AES-256 at rest on U.S.-based servers, isolated to each workspace, protected by access controls and never used for training, and that after cancellation data stays exportable for a limited period and is then permanently deleted. The Privacy Policy keeps personal information as long as necessary and names categories of service providers, including those that provide AI functions. No retention period, subprocessor list or incident notification commitment was located. Verified 22 September 2026.
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.
A published position with a higher cap for the risks that matter most, and output errors left with the firm. The Terms cap liability at twelve months' fees, but at three times that for breach of the no-training, confidentiality or data processing commitments, and let the customer recover the cost of investigating a breach and notifying clients, courts and regulators; the caps do not apply to gross negligence, wilful misconduct or fraud. Output accuracy is not warranted, and the customer indemnifies Concorda for claims arising from reliance on unverified output. Verified 22 September 2026.
Checked the home, FAQ, pricing, about and Trust Center pages and the Privacy Policy, and tried the /terms and /terms-of-service addresses, on 22 September 2026. No terms of service or customer agreement is published, so nothing states what Litmas stands behind when its output is wrong, its liability limits or any indemnity. Verified 22 September 2026.
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.
One named connection, undocumented. The home page describes automated patent contention charts with a USPTO connection. No document management, practice management, email or e-filing connection is named, and what the USPTO connection does is not described; content is exported in a standard format. Verified 22 September 2026.
Integration is claimed, not documented. The FAQ says Litmas AI can integrate with most case or practice management systems through APIs, naming Clio, Filevine and MyCase, and the Unlimited plan includes case management integration support; drafts export to Word, with PDF export described as coming. What connects, what syncs and how it is configured is not described. Verified 22 September 2026.
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.
Checked the home, Plans and About pages, the Terms and Data Processing Addendum and the Privacy Policy on 22 September 2026. No hosting provider, region, tenancy or deployment option is published; the addendum says the countries where data is processed are listed in a subprocessor list available on request. Verified 22 September 2026.
A single cloud service with its region stated. The Trust Center and FAQ say the platform is browser-based, that all data is stored on U.S.-based servers, and that each firm's data is isolated to its workspace on a privately housed server. Where AI processing takes place, and whether infrastructure is dedicated or shared, is not stated, and no other deployment option is offered. Verified 22 September 2026.
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.
A named attestation, with evidence on request. The home and Plans pages say Concorda is SOC 2 Type II audited, and the Data Processing Addendum commits to share current security documentation and any certifications or audit reports once a year on written request. A trust centre is linked from the site; its contents could not be viewed on the date shown. No auditor or report period is published. Verified 22 September 2026.
Named examinations, reports on request under NDA. The Trust Center says Litmas has completed a SOC 2 examination and an independent HIPAA compliance audit, and offers both reports to customers and prospective customers under NDA through a request form. The SOC 2 report type, auditor and period are not stated. Verified 22 September 2026.
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 architecture and its controls are described; the providers are not named. The Data Processing Addendum says Concorda sends limited customer data to third-party AI model providers only for features the customer uses, under enterprise or API terms that bar training and with zero or no retention, and gives 30 days' notice before adding or replacing any subprocessor, AI model providers included. Which providers and models are used is in a list available on request, not published. Verified 22 September 2026.
A proprietary layer is described without what sits underneath. The FAQ says answers are combined through Litmas AI's proprietary legal reasoning layer from the case file and verified U.S. court opinions, and the Privacy Policy lists service providers that provide AI functions. No model, model provider, inference location or change notification is published. Verified 22 September 2026.
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.
The self-serve price is published; Enterprise is quoted. The Plans page lists the Team plan at $350 per user per month, monthly or annual, with the first two weeks free and no demo or sales call needed, covering unlimited matters and uploads. Enterprise, for more than 10 users, has flexible pricing and custom terms on a call. Verified 22 September 2026.
A buyer can price it without a sales call. The pricing page publishes a per-case plan at $125 per case per month, pausable at any time, and an Unlimited plan at $999 per attorney per month with a free paralegal seat, personalised onboarding and pre-suit drafting; each case includes 5,000 pages, with additional pages at $0.01 each. A drafting service by Litmas's own legal team is priced at $1,499 per case per month per attorney. A free trial is offered. Verified 22 September 2026.
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.
Practice areas and work types are named, short of limits. The site addresses IP, plaintiff and commercial litigation teams and legal departments, lists the documents it drafts, from complaints and motions to interrogatory responses and contention charts, and says its case database covers every U.S. court. The courts, practice areas and document types it does not handle are not stated. Verified 22 September 2026.
Litigation work and practice areas are named, short of limits. The site covers solo practitioners to larger firms and in-house teams, names personal injury, commercial litigation, insurance defense and criminal defense, lists the motions and documents it drafts, and claims 50-state and federal case law coverage with local rule formatting. Where case law currency varies by court, the cadence is given only on request, and what it does not handle is not stated. Verified 22 September 2026.
The 12 legal signals, side by side
Recorded rather than graded. These are the questions a practitioner has to answer before a tool touches a client matter, and the answers are taken from public material only.
Client Data in Training
Can material a lawyer puts into this product be used to train a model?
The Terms say Concorda does not use customer content to train, retrain, fine-tune or otherwise adjust any AI model, for the customer or anyone else, and the Data Processing Addendum repeats the commitment and requires model providers to be configured so they cannot train on it. Breach of the no-training clause carries a liability cap three times the standard one. Aggregated, de-identified usage data may be derived but must not identify the customer, its clients or the substance of its content.
The Trust Center and FAQ state that workspace data remains private, is never used to retrain AI models, and is used only to power the firm's case analysis. No customer agreement is published, so the statement is not matched by a contractual term. Which model providers process case data is not stated.
Prompt and Output Retention
How long does the product keep what a lawyer typed, and can that be set to zero?
Customers can delete their content or ask in writing for it to be deleted, and content can be exported at any time and for 60 days after an account ends. The Data Processing Addendum then requires deletion within 30 days from production systems and within 90 days from backups, including by subprocessors, with written certification on request. Model providers are used with zero or no retention. A setting that keeps nothing inside Concorda is not described.
The Trust Center says that after cancellation data stays available for a limited retention period for export and is then permanently deleted, without stating the period. The Privacy Policy keeps personal information as long as necessary for its stated purposes. Nothing states how long prompts and AI outputs are kept during a subscription.
Ethical Walls and Matter Segregation
Does retrieval respect the firm’s ethical walls, or can the model read across them?
Checked the home, Plans and About pages and the Terms and Data Processing Addendum on 22 September 2026. The addendum describes role-based access controls within Concorda's own security program, and the terms let a customer cancel if the business passes to an adverse party; nothing describes separating matters or users inside a customer's account.
The Trust Center says each firm's data is isolated to its workspace and not shared with other users. It does not describe how access is controlled within a firm, whether cases can be walled off from particular users, or how isolation is enforced.
Third Party Request and Subpoena Notice
If someone subpoenas the vendor for a firm’s data, does the firm hear about it first?
The Data Processing Addendum commits Concorda, where legally permitted, to notify the customer promptly of a subpoena, court order or other legal process seeking its data, give it a chance to object or seek protection, disclose only what is legally required and cooperate in efforts to quash or narrow the demand. The Terms add advance notice where permitted. No transparency report was located.
The Privacy Policy says Litmas may disclose personal information in response to valid requests by courts or government agencies, or to comply with a legal obligation. Nothing addresses whether the firm will be told, and no customer agreement is published to address requests for case documents.
Primary Law Corpus Provenance
Where does the law in this product come from, and does the vendor have the right to use it?
The home page describes Concorda's own case database covering case law, statutes and local rules from every U.S. court, updated daily. Where the underlying opinions and rules come from, and the basis on which they are used, is not identified.
The FAQ and pricing page describe the corpus as verified U.S. court opinions with 50-state and federal coverage, updated on an ongoing basis at a frequency that varies by court and is given on request. The source of those opinions and the basis on which they are used are not identified.
Good Law Verification
Does the product tell you when the authority it just cited has been overruled?
Checked the home and Plans pages and the Terms on 22 September 2026. The site says every citation in the drafter is verified and research results are ranked by relevance; nothing addresses whether cited authority is checked for later negative treatment.
Checked the FAQ, pricing, how it works and home pages on 22 September 2026. Dual-layer validation is described as checking each citation for existence and accuracy against U.S. court opinions; nothing addresses whether cited authority is checked for later negative treatment, and the FAQ leaves weight and applicability to the lawyer.
Refusal and Uncertainty Behavior
What does the product do when the answer is not in the corpus?
Checked the home and Plans pages and the Terms on 22 September 2026. Nothing describes what drafting or research does when the record or the case law does not support a point.
The FAQ says that a citation which cannot be validated against a source is removed rather than shown, and the how-it-works page says the Evidence Mapper surfaces claims the record cannot support and shows where evidence is missing. No published evaluation shows that behavior.
Fabricated Citation Record
Does a public court record exist addressing fabricated or hallucinated legal citations in output from this product?
Searched the AI Hallucination Cases database maintained by Damien Charlotin on 22 September 2026 for Concorda, and no recorded case was returned. No court order, opinion or disciplinary record naming the product was located. This is a statement about the public record rather than a finding about the product.
Searched the AI Hallucination Cases database maintained by Damien Charlotin on 22 September 2026 for Litmas, and no recorded case was returned. No court order, opinion or disciplinary record naming the product was located. This is a statement about the public record rather than a finding about the product.
Bar Guidance Alignment
Has the vendor engaged in public with the ethics opinions its buyers are bound by?
The Terms leave the customer responsible for its applicable rules of professional conduct and require independent review of every output before filing. No ethics opinion is named.
The FAQ addresses professional responsibility in general terms: the filing is the attorney's, verification before filing is the lawyer's duty, the platform supports the duty to supervise, and attorneys have been sanctioned for AI-fabricated citations. No ethics opinion is named.
Billing and Fee Posture
Does the vendor address what happens to the bill when the work takes an hour instead of six?
The home page promises fewer late nights, a full day of drafting per party turned into deposition strategy time, and high-quality output without more headcount or outside spend. Nothing addresses how AI-assisted work is reflected in what a client is billed.
The FAQ says attorneys typically cut motion drafting time by 60 to 80 percent. Nothing addresses how AI-assisted work is reflected in what a client is billed or what a client should be told.
Outside Counsel Guideline Readiness
Can a firm get this vendor through a client’s AI clause without a bespoke negotiation?
The Data Processing Addendum says Concorda keeps a current list of subprocessors, including every AI model provider and the countries where data is processed, available on request, and gives 30 days' notice before adding or replacing one. Security documentation and audit reports are shared once a year on written request. The list itself is not published.
Checked the Trust Center, FAQ and Privacy Policy on 22 September 2026. The SOC 2 and HIPAA reports are offered under NDA on request; no subprocessor or model provider list and no client-facing disclosure material was located.
Court Disclosure Support
If a judge’s standing order requires an AI disclosure, can the product produce one?
Checked the home and Plans pages and the Terms on 22 September 2026. Nothing addresses recording or disclosing AI use in filings; the Terms leave review and disclosure to the customer.
Every fact and case reference in the output carries an inline pincite to its source, and the FAQ says the platform gives lawyers the traceability to show their work. No audit log, record of which model produced a draft, or export of AI use for disclosure to a court is described.
The questions both sides leave open
Derived from the records above rather than written, so it cannot favor either vendor. Take these into both conversations and ask each side the same question.
- AI Governance and Bias Disclosure
- Good Law Verification
Which one fits
Choose Concorda if
- You need privilege and training addressed in the agreement you sign. Concorda's terms bar using customer content to train, retrain or fine tune any model, state that privileged material and work product are processed only as a confidential service provider without waiver, and require model providers to run with zero or no retention.
- You want the vendor to carry more of the risk if it breaks its data promises. Concorda caps liability at twelve months' fees but at three times that for breach of its no training, confidentiality or data processing commitments, and lets you recover the cost of notifying clients, courts and regulators after a breach.
- You need deletion and breach notice on a clock. Concorda's addendum deletes content within 30 days from production and 90 days from backups with written certification, gives notice of a confirmed security incident within 72 hours, and gives 30 days' notice before any new subprocessor, AI model providers included.
Choose Litmas AI if
- You need citations you can check before a filing. Litmas AI checks every case, statute and pincite against verified U.S. court opinions through a dual layer validation, removes any citation it cannot validate rather than showing it, and links every remaining one inline so you can open it.
- You want to pay by the matter. Litmas AI charges $125 per case per month, pausable at any time, with 5,000 pages per case and extra pages at one cent, or $999 per attorney per month for unlimited cases with a free paralegal seat, and it offers a free trial.
- You want to see which elements your evidence proves. Litmas AI breaks each cause of action into its elements, maps the evidence in the record to each one to show what is supported and what is missing, and drafts discovery requests and deposition outlines from the gaps.
In summary
Concorda
Concorda, from Concorda, Inc., founded by former litigators, is an AI platform for IP, plaintiff and commercial litigation teams and legal departments, and is not the contract management company Concord. It drafts complaints, motions, briefs and discovery from the full matter record, builds patent contention charts, and searches case law, statutes and local rules across U.S. courts in its own database. The AI Legal Index grades it in the top two bands on ten of fifteen capability axes, with an A on AI centrality. Its terms bar training on customer content, address privileged material directly and commit to deletion timelines, and the Team plan is $350 per user per month. As of 22 September 2026 the index located no named customer, no accuracy measurement and no hosting location.
Litmas AI
Litmas AI, from Litmas, Inc., is a litigation platform for trial and appellate lawyers in personal injury, commercial, insurance defense and criminal defense practice. It answers case questions with pincites to the record and to U.S. court opinions, maps evidence to the elements of each cause of action, and drafts discovery, deposition outlines, motions and pleadings, removing any citation its dual layer validation cannot verify. The AI Legal Index grades it in the top two bands on nine of fifteen capability axes, with A grades on AI centrality and pricing: $125 per case per month or $999 per attorney. Data is stored on U.S. servers. As of 22 September 2026 the index located no published terms of service, no named model provider and no retention period.
Questions buyers ask
Concorda vs Litmas AI: which is better for drafting motions?
On published evidence Concorda sits in the top two bands on ten of fifteen AI Legal Index capability axes and Litmas AI on nine of fifteen. Concorda publishes a full agreement covering training, privilege, deletion, breach notice and liability. Litmas AI publishes no agreement but describes how it validates citations, names customers and prices by the case. Firms that must review vendor terms before use have more to read from Concorda.
How does Litmas AI check citations?
Litmas AI says every case, statute and pincite is checked against verified U.S. court opinions by a dual layer validation before the user sees it, that a citation that cannot be validated is removed rather than shown, and that each remaining citation links inline so the lawyer can judge its use. It does not say whether authority is checked for later negative treatment, and no error rate is published. Graded by AI Legal Index against 15 capability axes and 12 legal signals, including privilege handling and citation accuracy, from each vendor's own published materials, verified September 25, 2026. No vendor pays for placement.
Does Concorda train AI on client documents?
No, by contract. Concorda's terms state that it does not use customer content to train, retrain, fine tune or otherwise adjust any AI model, its data processing addendum requires model providers to be configured so they cannot train on it, and a breach of that clause carries a liability cap three times the standard one. Anonymized usage data may be derived only if it cannot identify the customer, its clients or the substance of its content. Graded by AI Legal Index against 15 capability axes and 12 legal signals, including privilege handling and citation accuracy, from each vendor's own published materials, verified September 25, 2026. No vendor pays for placement.
How much do Concorda and Litmas AI cost?
Concorda's Team plan is $350 per user per month with two weeks free and no sales call, and Enterprise is quoted for more than ten users. Litmas AI charges $125 per case per month or $999 per attorney per month for unlimited cases, with 5,000 pages per case and extra pages at one cent each, plus a separate drafting service by its own legal team at $1,499 per case per month. Graded by AI Legal Index against 15 capability axes and 12 legal signals, including privilege handling and citation accuracy, from each vendor's own published materials, verified September 25, 2026. No vendor pays for placement.
What do Concorda and Litmas AI both leave unpublished?
Which models do the drafting, and how often they are wrong. Neither names its model providers in public, and neither publishes an accuracy measurement or says whether cited cases are checked for later negative treatment. Neither publishes an AI governance position, describes walls between matters inside a firm, or offers a record of AI use a lawyer could show a court. Neither addresses how saved drafting time should appear on a client's bill. Graded by AI Legal Index against 15 capability axes and 12 legal signals, including privilege handling and citation accuracy, from each vendor's own published materials, verified September 25, 2026. No vendor pays for placement.
Two readings to weigh. Litmas AI publishes no terms of service, customer agreement or data processing addendum, so its commitments against training and on isolation are policy statements on its trust center and FAQ, and nothing published states its liability or what it stands behind when output is wrong. Concorda names no customer anywhere, and its subprocessor list, including AI model providers and processing countries, is available only on request. Neither publishes an accuracy measurement for its citation checks. Both records were verified on 22 September 2026. Neither vendor reviewed this page.
Neither vendor paid for inclusion, placement or a grade, and neither reviewed this page before it published. Everything above comes from public material on the dates shown. How the index grades.