Z
Zuva

Zuva is contract analysis AI from the team that built Kira Systems. When Kira was sold to Litera in 2021, its founders kept a copy of the underlying AI and about 30 of its staff, and that became Zuva Inc. of Toronto. Its extraction models recognize more than 1,400 contract provisions and classify more than 225 document types, and lawyers have trained them since 2011 rather than building on a general purpose language model.

Zuva sells the technology five ways. Zuva Analyze is a self serve review tool at $10 a document that exports to Excel or Word, and its Answers fields pass extracted clauses to a language model that picks from fixed answer choices. The Zuva API puts extraction, classification and language detection behind a REST interface, with a Power Automate connector that Microsoft's legal team has used. Zuva Create lets lawyers train their own fields without code.

Diligence Reviews pair the AI with former large firm M&A lawyers to produce sell side pre diligence reports and disclosure schedules, mostly for $10,000 to $75,000. Bidder Question Management, in free beta, tracks buyer questions in an auction. The service runs on Microsoft Azure in the United States or Europe or self hosted, and Zuva states SOC 2 Type II certification. Under the Kira sale, Zuva does not sell interface products directly to law firms until later in 2026.

Compared with

Zuva, head to head

Side by side on the same 15 capability axes and 12 legal signals: where the two part company, and what neither one publishes.

Build your own comparison

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.

AA on AI CentralityThe artificial intelligence is the product. Remove the models and there is nothing left to sell.

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.

Machine learning is the product in every form Zuva sells. The API is a REST service that extracts structured data from contracts, classifies documents and detects their language, and Zuva Analyze puts the same extraction behind a self serve interface with workflows by document type. Zuva Create lets a customer train new fields on its own documents, and Answers fields send extracted clauses to a language model that picks from fixed answer choices such as consent required or freely allowed.

Diligence Reviews, the service line, start with the AI scanning every contract before a lawyer reviews the findings. Bidder Question Management applies the technology to spotting similar buyer questions across bidders in an auction. Remove the models and nothing is left. Zuva has no document management, workflow or storage product underneath, and its security overview says the API is not meant as a long term repository.

Source: Vendor Published
BB on Citation Accuracy and Hallucination DisclosureGrounding is real and documented, with linked primary sources and a described retrieval method, short of published accuracy figures an outsider can test.

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.

Zuva's extraction returns the contract's own text. The extraction documentation describes each result as an instance of the requested text in the document, and Analyze lets a reviewer jump from a result to the clause it came from. The API page says the same input gives the same extraction every time and describes the models as biased toward finding everything relevant. Answers fields limit the language model to a fixed set of answer choices, with evidence from the document.

Zuva gives no recall, precision or error rate for its fields. The Analyze page says other systems give maybe 70 to 80 percent accurate results and that lawyers using Analyze matched manual accuracy at 40 to 80 seconds a page, without a test set or figure. A February 2026 study with the general counsel of JWX compared Zuva's output with law firm diligence memos on two completed deals. On the larger deal Zuva missed one change of control item the law firm found, and the law firm missed two that Zuva found.

Zuva's Reviews FAQ says the service may miss things. The out of the box fields were trained mainly on publicly available English documents such as EDGAR filings.

Source: Vendor Published
BB on Autonomy and Oversight ModelA written commitment that the models work alongside a supervising lawyer, with real review surfaces, short of the full control structure: commonly the threshold at which the system stops or what happens after it is wrong.

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.

A lawyer is built into the service, while review of the software is left to the customer. In Diligence Reviews the AI scans every contract, then an experienced corporate lawyer reviews and contextualizes the findings before a pre diligence report is delivered, and disclosure schedules go to the seller's counsel. In the February 2026 JWX study, a Zuva reviewer with large firm M&A experience validated the AI output, mainly by removing false positives.

Zuva Analyze has an interface built for verification, where each result links to its clause. In its featured case study, a motorsports team used paralegals for first pass review and counsel for second line checks. The API leaves oversight to the developer building on it. The product is extractive rather than agentic. It returns text found in the document and does not draft, send or decide anything, and the same input gives the same extraction every time.

Zuva describes no confidence measure for an extraction and no check for a provision the models missed, and its Reviews FAQ concedes that the service may miss things, as law firms do.

Source: Vendor Published
AA on Operational and Outcome EvidenceNamed firms or legal departments, dated, with figures for what changed and a method a reader can assess.

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.

Zuva backs its claims with a dated benchmark against law firm work. In a February 2026 study, David Pashman, general counsel of JWX, gave Zuva the data rooms of two completed acquisitions. Zuva reviewed them blind for change of control, restrictive covenants, most favored nation and first refusal rights, and the two then reconciled its results against the law firms' memos issue by issue. On the larger deal Zuva found seven non solicitation restrictions the firm did not list and missed one change of control item the firm found.

The study says it covers a subset of issues and is not evidence that Zuva can replace full scope diligence. Zuva also reports that on a private equity deal, two Chambers Band 1 firms found no contract it had missed in the disclosure schedules, with that case study under NDA. Microsoft's legal team cut time on routine contract requests by roughly 75 percent with the API in Power Automate, without a date or method. Jim Miller, general counsel of Clairvest, is quoted on Reviews, and an unnamed motorsports team reviewed 2,000 contracts in Analyze.

Source: Vendor Published
BB on Privilege and Confidentiality PostureSubstantive published commitments on confidentiality and training use, short of the full picture: commonly silence on segregation between users or matters, or on what the underlying model provider may retain.

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.

Ownership of customer material, and how Zuva may use it, is set out in the DocAI and AI Trainer License Terms. The customer owns its documents, customer data and outputs (section 6), and Zuva treats them as confidential information held to at least a commercially reasonable standard of care (section 8). Section 2(g) confines Zuva's improvement right to usage data, defined as information about access to and use of the service that excludes customer data.

Fields a customer builds are for its exclusive use and are deleted when the agreement ends. The security overview says no other customer, and not Zuva itself, may use a customer's fields and data without permission, and the Create page says training documents are never used to train shared models. Model storage does not rely on the content of the training documents, and only the creating token can reach an account's data.

The agreement contains no express training prohibition. Zuva does not address privilege or work product, or what the language model provider behind Answers may keep.

Source: Vendor Published
CC on UPL and Professional Responsibility PostureA boilerplate disclaimer sits in the terms while the marketing describes the product in advice terms, or the intended audience is left ambiguous.

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.

Diligence Reviews are sold in place of law firm diligence work, to buyers who are not all lawyers. The home page and the Reviews pricing page offer sell side pre diligence and disclosure schedules with lawyer review. The buyers named are chief financial officers, general counsel and private equity firms, and the price is about 80 percent below large firm fees. Zuva prepares disclosure schedules in coordination with the seller's law firm, formats them to that firm's conventions and describes them as ready for counsel review.

Its Reviews FAQ says that in United States and Canadian deals, liability for disclosure errors tends to sit with the buyer. Zuva does not say whether the work is legal advice, whether Zuva practices law, or whether a lawyer and client relationship arises. The accuracy disclaimer in the license terms limits output rather than drawing the advice line. Analyze is addressed to legal teams and the API to developers. Under the Kira sale, Zuva will not sell interface based contract analysis directly to law firms until later in 2026, and in the meantime it takes referrals of sell side clients from law firms.

Source: Vendor Published
CC on AI Governance and Bias DisclosureResponsible AI principles are published without a mechanism, a testing regime, or anything a buyer could audit.

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.

Training data sources are stated, along with research on how the models work. Legal knowledge engineers trained the out of the box fields mainly on publicly available English documents such as EDGAR filings, with a subset in German and Brazilian Portuguese. Zuva says many of the lawyers who trained them came from firms including Goodmans, Sidley, Allen & Overy, Stikeman and Torys. The science page lists papers by Zuva's team on due diligence retrieval, document viewers, clustering and how lawyers work with AI.

Adam Roegiest, vice president of research and technology, leads research, development and product. A March 2026 post by chief executive Noah Waisberg sets out four levels of testing legal AI, from internal benchmarks to tests against real lawyer work product. Zuva states no responsible AI policy and describes no testing a field or model passes before release. It reports nothing on how extraction quality varies across document types, languages or drafting styles. Custom fields are trained on the customer's own documents.

Source: Vendor Published
BB on AI Safety and Data StewardshipSubstantive published policy covering most of the ground, short of the full set: commonly no named subprocessor list or no stated incident practice.

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.

Retention, deletion, encryption and segregation are set out in Zuva's security overview and license terms. Uploaded documents expire after seven days by default, the expiry can be set from one hour to fourteen days, and files deleted early through the Files API cannot be restored. The API page offers retention down to none. Each account's data is reachable only by the token that created it, sits in a client specific object store on a 256 bit encrypted disk, and travels over TLS 1.2.

Zuva's systems team has administrative access to production in every region, and senior engineers can read application logs only with management approval. All staff pass background checks and are trained on client confidentiality and insider trading. Section 8(d) of the license terms commits Zuva to physical, managerial and technical safeguards and refers to a data protection addendum. Zuva names no subprocessors and states no incident notification period, and its privacy policy covers registration and site tracking rather than uploaded documents.

Source: Vendor Published
BB on AI Liability and RecourseA real published position on liability, short of the full picture: commonly a stated indemnity without scope or caps.

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.

Zuva's license terms set a liability cap with carve outs and an intellectual property indemnity, and disclaim the accuracy of results. The DocAI and AI Trainer License Terms and Conditions, last updated 11 May 2022, are governed by Ontario law. Section 7(a) commits Zuva to indemnify the customer against third party claims that the services, used as the agreement allows, infringe intellectual property rights. The remedy is a license to continue, a replacement, or termination with a refund of prepaid fees.

Section 10 caps each party's liability at the amount paid in the six months before the claim, and carves fraud, willful misconduct and intellectual property infringement out of the cap. Section 9(b) disclaims all other representations and warranties, including any warranty that results will be accurate or reliable. These terms cover the API and the software. Zuva posts no engagement terms for Diligence Reviews and does not say what it stands behind when a disclosure schedule prepared with its lawyers is wrong.

Its Reviews FAQ says that in United States and Canadian deals, liability for disclosure errors tends to sit with the buyer, especially where the deal carries representations and warranties insurance.

Source: Vendor Published
BB on Practice Systems Integration DepthReal integrations exist and are documented, short of depth: named connections without a description of what they actually move.

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.

API documentation is open to read and written for developers. It covers file submission, OCR, classification, language detection, field extraction and custom field training, with a quick start, Postman testing, a Python SDK tutorial and a full API reference. A Power Automate connector runs reviews when a file reaches a SharePoint folder, an email arrives or a form is submitted, and Microsoft's legal team built a triage flow on it.

Zuva's October 2024 post on renaming its products names LawVu and Kofax TotalAgility among tools combined with its API. Analyze exports to Excel, including one sheet across all documents, and to Word. Bidder Question Management imports Word and Excel request lists and exports to Excel in a format set to the user's own template. Zuva offers no connector to iManage, NetDocuments or a matter management system, so deeper integration is engineering by the customer or a partner.

Source: Vendor Published
BB on Deployment Model and Data ResidencyDeployment model is stated clearly with partial residency detail, or residency is offered without the processing location being addressed, or the tenancy model is stated on its own with no residency detail published.

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.

Shared cloud, private cloud and self hosted deployments are all available. The API servers run on Microsoft Azure in the United States and Europe, and the customer picks the region when it creates a token. A private cloud in another Azure region, or self hosting on Azure or AWS, costs extra, and Zuva will deploy in other environments for large enterprise contracts. The security overview says customers that must keep data inside their own IT boundaries use the on premises option, so contracts never leave their systems.

Section 2(d) of the license terms places hosting in the jurisdiction the customer selects and user authentication in Canada or the European Union. The Analyze enterprise plan lists hosting options. The API page says its models are small and data efficient enough to self host on Azure or AWS out of the box. Zuva does not say where the language model behind Answers and summaries runs, or whether it stays in the selected region.

Source: Vendor Published
BB on Security Certifications and Trust CenterCertification is real and stated, short of accessible evidence: a named standard without scope, date, or a way to obtain the report.

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.

SOC 2 Type II certification is stated on Zuva's home page, product pages and API security overview, and the overview says the Zuva API has obtained SOC 2 certification. Zuva names no auditor, report period or scope, and links no trust portal or route to request the report. The overview describes the controls in detail, including TLS 1.2 in transit, 256 bit encrypted storage, a disaster recovery policy reviewed at least once a year, continual vulnerability testing and source code security reviews.

Zuva discusses the results of its latest third party security testing with customers under NDA. It does no outsourced programming and runs on Microsoft Azure, which holds its own SOC 2 Type II and ISO 27001 certifications. Named policies cover incident management, information security, privacy, secure software development and change management.

Source: Vendor Published
BB on Model Supply Chain DisclosureThe supply chain is partly disclosed: providers named without change notification, or architecture described without the providers.

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 extraction models are Zuva's own, carried over from Kira Systems at the 2021 sale and improved since. Its research team works on the underlying models and on wrappers around general purpose AI. Answers fields pass the extracted clauses to a language model that selects from fixed answer choices, and Zuva's September 2024 post on ABA Formal Opinion 512 says this feature does not self learn. Zuva does not name the model behind Answers or the summaries in Analyze.

A September 2026 post says that when a diligence project needs summaries, Zuva runs documents through GPT or Claude and uses Analyze to check the results. On a buy side deal that summer, OpenAI's most advanced GPT model did the first pass. OCR runs on Kofax OmniPage. Zuva commits to no notice when a model or provider changes, and custom field training runs under the DocAI and AI Trainer License Terms.

Source: Vendor Published
BB on Commercial TransparencyReal pricing is published for part of the range, with enterprise tiers withheld, or the unit and structure are stated without the figure.

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.

Prices for the self serve products and the review service sit on Zuva's own pages. Analyze costs $10 per document with a $5,000 minimum, five free documents to try and volume discounts. The API is free for 25 files, classifications, language checks and extractions a day, then costs $1.25 a document plus $0.0015 per field per page, with a $5,000 minimum. Most Diligence Reviews deals cost $10,000 to $75,000 at a fixed fee for an agreed scope, with large or complex scopes from $125,000 and quotes within 48 hours.

Bidder Question Management is free during its beta and will then be priced per deal, like a data room. Enterprise volume, private cloud, self hosting and rush delivery are quoted. From these figures, the cost of a review project, a batch of contracts or an integration can be worked out before contacting Zuva. The one open item in a fixed fee review is disclosure schedule work, which Zuva says can change late in a deal.

Source: Vendor Published
AA on Firm and Practice CoverageWho the product serves is documented precisely: firm segments, in house and government use, and the practice areas actually supported, with the limits stated.

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.

Three kinds of buyer use Zuva. Diligence Reviews and Bidder Question Management serve the sell side of M&A, for chief financial officers, general counsel, private equity firms selling portfolio companies and investment banks running auctions. Reviews cover active agreement confirmation, a red flags review, disclosure schedule preparation, employment agreement IP review and benefit plan cataloging, and Zuva has also run buy side summary work.

Zuva Analyze is for legal teams and in house lawyers working on M&A, entity restructuring, NDA portfolios, CLM population, lease abstraction and compliance. The API and Create are for developers and product teams building contract automation. The field library covers more than 1,400 provisions and more than 225 document types in three levels of classification. Zuva states three limits. Under the Kira sale it does not sell interface based contract analysis directly to law firms until later in 2026.

The out of the box fields are mainly English, with a subset in German and Brazilian Portuguese and training available in other Latin script languages. The API cannot read handwriting or zip files.

Source: Vendor Published
Sources on file

10 public documents

The public pages on file for Zuva, with the recorded signals each one supports and the date it was last read. Open any of them and check the reading against the record.

Pricing

$10 per document

USD, as published, never converted
  • Zuva charges by the document.
  • Its review tool costs $10 for each contract, with at least $5,000 spent.
  • Developers can use the API free for small amounts, then pay per document.
  • A full sale diligence review by Zuva's lawyers costs $10,000 to $75,000 for most deals.
  • Its tool for tracking buyer questions in a sale is free while in beta.
  • Big company plans and private hosting are priced on request.

Zuva Analyze, the self serve review tool for legal teams, costs $10 a document pay as you go, with a $5,000 minimum commitment, five documents free to try and enterprise volume pricing on request. The Zuva API has a free tier of 25 files, classifications, language detections and extraction requests a day, with up to ten fields per request. Beyond that it charges $1.25 a document plus $0.0015 per field per page, with a $5,000 minimum and bulk discounts, and an enterprise tier at custom pricing with service levels.

Diligence Reviews are mostly fixed fee for an agreed scope, at $10,000 to $75,000 for about three quarters of deals and $125,000 or more for large or complex scopes, quoted within 48 hours. They take one to two weeks, with rush delivery at extra cost. Bidder Question Management is free during its beta and will then be priced per deal by deal size and engagement length.

Implementation: None stated for the cloud products. Private cloud and self hosting are offered for an additional fee that is not priced.

Confidentiality and data terms: Not applicable. The license terms provide a data protection addendum where data protection law requires one.

Note: The $5,000 minimum applies to both pay as you go plans. Enterprise volume rates, private cloud and self hosting fees, and the rush premium for reviews are quoted per customer. The $10 figure is Analyze, the product a legal team buys. The API's per document rate is lower and adds per field charges.

Legal Signals

What each signal means

A 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.

Confidentiality and Privilege

Client Data in Training

Can material a lawyer puts into this product be used to train a model?

Never, in policy only

A public policy or trust page states no training on customer content, with no matching term located in the published agreement.

Zuva's product pages commit that customer data does not train other customers' or shared models, and its license terms do not mention training. The security overview says no other customer, and not Zuva itself, may use a customer's fields and data without permission. The Create page says training documents are never shared with other customers or used to train shared models. Zuva's post on ABA Formal Opinion 512 says Answers, its language model feature, does not self learn.

The license terms, last updated 11 May 2022, confine Zuva's improvement right to usage data about access to the service, which excludes customer data. The standard fields were trained on public filings.

Source: Vendor Publishednever shared with other customers or used to train shared modelsAs of Oct 10, 2026Evidence

Prompt and Output Retention

How long does the product keep what a lawyer typed, and can that be set to zero?

Customer set, zero available

The customer sets the retention window and no retention is an available setting.

The customer sets retention, and keeping nothing is offered. The API page says documents can be deleted immediately after processing, kept for a set period or anything in between. The security overview says uploads expire after seven days by default, and the expiry can be set from one hour to fourteen days, with extensions while processing continues. Files deleted early through the Files API cannot be restored. Analyze deletes documents when the customer chooses, and Diligence Reviews delete them on request.

Fields a customer trains are deleted from its account when the agreement ends. Zuva does not say how long the language model behind Answers keeps the clauses it receives.

Source: Vendor PublishedYou control retention: no retention, a week, whatever you needAs of Oct 10, 2026Evidence

Ethical Walls and Matter Segregation

Does retrieval respect the firm’s ethical walls, or can the model read across them?

Own model, documented

The product maintains its own permission model, documented, requiring the firm to keep it aligned.

Data separation happens at the level of the API token. The security overview says all data in an API client account is stored and permissioned so that only the token that created it can reach it, in a client specific object store. A customer can create several tokens from the admin console, each limited to the files and requests it created, with usage tracked per token, so projects run under separate tokens stay apart.

Custom fields stay private to the customer's organization. Zuva does not describe walls between matters inside Analyze or any link to document management permissions.

Source: Vendor Publishedonly the token used to create the data can access itAs of Oct 10, 2026Evidence

Third Party Request and Subpoena Notice

If someone subpoenas the vendor for a firm’s data, does the firm hear about it first?

Notice committed

Terms commit to notice where lawfully permitted. No transparency report located.

Section 8(c) of the license terms commits Zuva to give prompt written notice of a compelled disclosure where the law permits. It covers subpoenas, civil investigative demands and requests in legal proceedings for confidential information, which includes the customer's documents and data, and requires reasonable efforts to preserve confidentiality and cooperation with a protective order. Section 2(d) places hosting in the jurisdiction the customer selects, and self hosting keeps documents inside the customer's own systems. Zuva Inc. is a Toronto company running on Microsoft Azure. Zuva issues no transparency report on requests it receives.

Source: Vendor Publishedshall, if permitted under Applicable Laws, provide Disclosing Party with prompt written noticeAs of Oct 10, 2026Evidence
Accuracy and Authority

Primary Law Corpus Provenance

Where does the law in this product come from, and does the vendor have the right to use it?

Sources named, basis unstated

Sources are identified without stating the license or rights basis.

Zuva's fields work on contracts the customer supplies rather than on primary law. The API FAQ says Zuva's legal knowledge engineers train the out of the box fields on repositories of publicly available documents such as EDGAR, primarily in English, with a subset in German and Brazilian Portuguese. The API page says the fields took tens of thousands of hours of work by experienced lawyers, many from firms including Goodmans, Sidley, Allen & Overy, Stikeman and Torys.

Zuva names no collection beyond EDGAR and states no rights basis or update cadence. Custom fields are trained on the customer's own documents, and service partners have built further fields that they share.

Source: Vendor Publisheduse repositories of publicly-available documents, such as EDGAR, to train AI fieldsAs of Oct 10, 2026Evidence

Good Law Verification

Does the product tell you when the authority it just cited has been overruled?

Not addressed

No located public material addresses whether authority is checked for subsequent history.

Provision extraction and document classification are Zuva's work, with no case law or statutes retrieved or cited, so a check of subsequent history sits outside the product. Its fields return the contract's own text rather than legal authority, and Answers fields choose from fixed categories such as consent required or freely allowed. Analyze is pitched for compliance audits when regulations change, but Zuva describes no way to flag a provision whose effect depends on a rule that has since changed. Contract Central, its clause glossary, explains provisions such as indemnification and force majeure in general terms.

Source: Operator VerifiedAs of Oct 10, 2026Evidence

Refusal and Uncertainty Behavior

What does the product do when the answer is not in the corpus?

Documented

The vendor describes refusal or abstention behavior in public materials.

The extraction returns nothing for a field it does not find. The field extraction documentation says each requested field comes back with zero or more extractions, each an instance of the text in the document. The API page says the same input gives the same extraction every time. Answers fields restrict the language model to a fixed set of choices with evidence from the document, such as consent required, notice required or freely allowed.

Zuva describes no confidence measure for an extraction and no abstention option for Answers or the summaries in Analyze. The API FAQ states that handwritten documents cannot be read.

Source: Vendor Publishedthere is an array of zero or more extractionsAs of Oct 10, 2026Evidence

Fabricated Citation Record

Does a public court record exist addressing fabricated or hallucinated legal citations in output from this product?

None located

No court order, opinion or disciplinary record addressing fabricated or hallucinated legal citations produced by this product has been located as of the date shown. This is a statement about the public record on that one subject, not a finding about the product, and this signal is not a litigation history.

The AI Hallucination Cases database maintained by Damien Charlotin, which records court decisions worldwide that address hallucinated AI content and the tool involved where known, has no entry naming Zuva or Zuva Analyze. Zuva extracts contract text for diligence and portfolio review rather than drafting court filings, so its output does not ordinarily reach a brief. Its September 2026 post on summarizing customer contracts reports that GPT invented terms in diligence summaries, such as a 15 day notice period the contract did not contain. Zuva's lawyers found them by reading the summaries against the agreements.

Source: Operator VerifiedAs of Oct 10, 2026Evidence
Professional Responsibility

Bar Guidance Alignment

Has the vendor engaged in public with the ethics opinions its buyers are bound by?

Named guidance addressed

Public materials engage with at least one named ethics opinion.

Zuva's September 2024 post by chief executive Noah Waisberg and research head Adam Roegiest analyzes ABA Formal Opinion 512. It reads the confidentiality section as applying to United States lawyers at law firms who put client information into self learning generative AI, and sets out how to test a tool against it. The post says Zuva's Answers feature uses generative AI to choose from fixed answers and does not self learn, and that under the opinion's terms a tool that is not generative AI falls outside it.

No product surface maps Zuva to guidance by jurisdiction, and Diligence Reviews, delivered with former large firm M&A lawyers, carry no statement on professional duties.

Source: Vendor PublishedThis Answers technology uses generative AI, but it does not self-learn.As of Oct 10, 2026Evidence

Billing and Fee Posture

Does the vendor address what happens to the bill when the work takes an hour instead of six?

Outside the fee relationship

The product does not touch a fee between a lawyer and a client. It operates before an engagement exists, or it is bought by a team that bills no client for the work. Savings claims aimed at the buyer’s own cost are recorded in the summary and do not make the row a savings claim, because no client bill is in the loop.

In house legal teams, developers, companies selling a business and their investment banks buy Zuva's products, and none of them bills a client for the work. Under the Kira sale, the interface products are not yet sold to law firms, which refer sell side clients instead. Diligence Reviews are priced at a fixed fee for an agreed scope with no hourly billing, about 80 percent below large firm fees, which Zuva puts at $300,000 to $700,000 for sell side diligence.

The saving is aimed at the seller's own cost and at lighter law firm work at signing. Zuva says nothing about how a firm that bills a client should treat AI assisted work.

Source: Vendor PublishedAs of Oct 10, 2026Evidence

Outside Counsel Guideline Readiness

Can a firm get this vendor through a client’s AI clause without a bespoke negotiation?

On request only

The material exists behind a sales conversation or an executed agreement.

Azure hosting in the United States or Europe, customer set retention, token level segregation and encryption are described in Zuva's security overview, readable without an agreement. It also covers staff background checks and limits on administrative access, and Zuva states SOC 2 Type II certification. Zuva shares the results of its latest third party security testing under NDA. The license terms commit Zuva to provide the security overview on request and a data protection addendum where data protection law requires one.

Zuva names no subprocessors and does not name the language model behind Answers, so that detail comes through the agreement or a security review.

Source: Vendor PublishedAs of Oct 10, 2026Evidence

Court Disclosure Support

If a judge’s standing order requires an AI disclosure, can the product produce one?

Not addressed

No located public material addresses court disclosure or verification certification.

Contract provisions extracted for diligence and portfolio review export to Excel or Word, and the output is not a court filing. Analyze links each result to its clause, and Diligence Reviews deliver a report and disclosure schedule sections formatted to the seller's law firm conventions. Zuva describes no record of which model produced an extraction or an answer, which results a person reviewed, or how its lawyers sign off their work.

Uploaded documents expire after seven days by default unless the customer sets another period. Zuva offers no disclosure template.

Source: Operator VerifiedAs of Oct 10, 2026Evidence
Contact

Correct a record, or ask how something was graded

Every grade and every signal on this index is drawn from public sources and dated. If a record is wrong, out of date, or missing an artifact the index did not locate, send the source and it will be reviewed and the record redated. Vendors are welcome to submit documentation. Nothing on this index is for sale, including a listing, a placement, or a grade.

AI Legal Index

The AI Legal Index is an independent index that tracks changes to AI vendors in legal. It holds 465 vendors across 9 categories, each graded on the same 15 capability axes and recorded against 12 legal signals, from public sources. No vendor pays for inclusion, placement, or rating.

Index Status
Last index update
November 10, 2026
The AI Legal Index is an editorial reference. It is not a regulatory body, not a law firm, and nothing published here is legal advice or a recommendation to retain or avoid a vendor. Records are verified against published sources, bar guidance and public court records. Where a record reads not addressed, the material was not located in public sources on the date shown. See the Methodology page for evaluation standards and limitations.
© 2026 AI Legal Index
3801 N Capital of Texas Hwy, Ste E240 · Austin, TX 78746