LawVu vs Xakia: how they compare in 2026
LawVu and Xakia both sell a legal operations platform to in house teams, covering intake, matters, contracts, spend and reporting. LawVu, from Tauranga, New Zealand, goes deeper on Microsoft, with LawVu Draft in Word, a Power Platform connector and an MCP server. Its Assistant for the legal team and a second assistant for business users both ask before they change records. Xakia, from Melbourne, targets teams of 2 to 200 and lists its prices, from $100 to $230 per user a month, month to month. Its AI reviews contracts in Word against the customer's playbook and extracts key terms with a confidence score and the source clause. Both publish their customer terms and AI terms, both let customers choose a data region, and both bar their AI providers from training on customer inputs. Xakia's addendum also lets it train its own AI features on anonymized data derived from those inputs. LawVu has held ISO 27001 since 2019, and Xakia holds ISO 27001:2022 and SOC 2 Type 2.
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.
Matter records, intake forms, the contract repository and invoice approvals all run in LawVu without generative AI. Intelligence adds extraction, billing guideline review, answers and record changes to that foundation. The Assistant replaces the earlier Workspace Intelligence chat and works across the spend module as well as matters and contracts. Its invitation beta and the Business Portal proxy preview have narrower availability than the general LegalOS marketing. Lens is an early access contract analysis feature, with Data Protection and Confidentiality available and other category templates still planned. Draft supplies Word redlining and clause reuse. Administrators can build AI Workflows that classify new matters, route contracts or create tasks. Manual and new-record triggers are supported; scheduled triggers remain forthcoming. AI reaches several operational jobs, while the underlying system of records remains useful without it.
Xakia's AI powers contract review, term extraction, summaries, plain language retrieval and invoice review within a broader legal operations platform. Contract reviewers apply a chosen playbook in Word and select Pragmatic, Balanced or Strict review settings. Extracted terms retain the source clause and require approval before entering the repository. Invoice analysis applies billing guidelines and flags inconsistencies in totals, tax and vendor details. AI remains optional, with extra charges on Advance and Professional and inclusion on Enterprise and All In. The underlying matters, tasks, intake, spend and entity workflows can operate without models. Connections from Claude and Copilot remain forthcoming, initially with read access only.
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.
Lens lets a reviewer expand an extracted clause and open its source contract in context. It accepts DOCX and PDF files, with citation highlighting currently available only for PDFs. A sample run can cover up to 25 contracts before a full extraction. Its grid separates Not Found, where no relevant provision is identified, from Unclear, where conflicting or ambiguous language cannot be mapped confidently to a value. The Business Portal assistant can supply source excerpts and direct unsupported questions into a legal request. Draft compares proposed wording with the customer's clauses and playbooks. These controls help a person check an answer, which is a different matter from a measured accuracy rate. The AI service terms require human review and disclaim completeness and accuracy. No measured legal error benchmark accompanies these features.
Xakia's AI points back to the customer's own documents. Key terms extraction returns each term with a confidence indication, the clause it came from and the exact text it relied on. A user reviews each value, corrects it and clicks Apply before anything reaches the contract repository. Contract review marks up a third party contract in Word against the customer's chosen playbook, governing law, market practice and basic drafting standards, at a Pragmatic, Balanced or Strict setting. It flags each deviation with a suggested redline for the lawyer to review. The query generated from a plain English request remains visible, editable and reusable. The AI addendum says output may not be accurate or up to date and must be reviewed by a lawyer. The FAQ commits to expanding AI without compromising accuracy, and Xakia gives no accuracy figure, error rate or test results.
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 conversational Assistant asks for approval before creating or updating records and groups batch operations into at most ten actions per approval. Changes and declined approvals enter its audit history. AI Workflows are different: an administrator configures actions, conditions and optional approval steps, then activates the workflow for manual or event-driven execution. The editor has test mode, version restoration and step-by-step execution results; scheduling is still forthcoming. MCP-connected external tools can create matters, assign tasks and trigger contract workflows. LawVu recommends human confirmation for their writes but does not describe a mandatory approval on every MCP action. Existing user permissions constrain each connection. The acceptable use policy bars consequential automated decisions without oversight, testing or prior risk evaluation, while product-specific workflow settings determine when a person intervenes.
Every Xakia AI feature requires a human decision before material is saved, applied or sent. Contract review proposes revisions inside Word against the selected playbook, and a reviewer decides which changes to use. Extraction exposes each value and its supporting clause before the user presses Apply. Invoice reviewers can waive findings, and their decisions enter an audit trail. The AI addendum requires legal practitioner supervision and prohibits consequential automated decisions without proper human review. These are usable review mechanisms, but the service does not specify a general stopping rule for unsupported reasoning or a complete correction and escalation process after an erroneous result.
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.
Sizzling Platter corporate counsel Samuel Gray attributes a three-month implementation to LawVu's migration support, including the bulk creation of more than 700 matters from spreadsheets. The customer story estimates that avoiding manual entry saved up to 175 hours. It also describes faster access to matter context, links between stores and acquisition agreements, SharePoint collaboration and a Mexico team's contract creation wizard. Its estimate of ten to fifteen hours saved each week depends on assumed daily matter activity, rather than a controlled study. The story has no visible publication date and does not isolate AI's contribution. Named users at Axel Springer and AES also describe Word adoption and shared drafting knowledge. These are identifiable customer accounts of operating benefits, without independent measurements of the newer Assistant or Lens.
Coca-Cola Europacific Partners uses Xakia to consolidate the Australian legal team's matters and replace paper forms and monthly spreadsheet reporting. Its deputy general counsel describes reporting delays of up to two months before adoption and current visibility into matters afterward. The team also used the information to identify trademark workload and allocate that work to an internal specialist. Other named users describe workload allocation, request handling and integrated files, including Contact Energy, Kiwi Property and the City of Murfreesboro. These are operational accounts from identifiable organizations. They do not measure the contribution of current AI features or provide a controlled before and after study of contract or invoice analysis.
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.
LawVu's AI service terms prohibit third parties from using customer data or outputs to improve or train AI models. Its AI FAQ extends the no-training position to LawVu itself and says Azure OpenAI monitoring is disabled under an exemption. Restricted matters admit named members, and the native Assistant and MCP server enforce the requesting user's existing permissions. A significant history limitation remains: a conversation is not removed when a user later loses access to the matter discussed in it. Organization administrators can review Business Portal conversations. MCP also transmits relevant legal data to the customer's chosen AI provider, whose processing terms are outside LawVu's control. The SaaS confidentiality clause protects customer information but permits necessary disclosures and professional advisers; it does not promise that uploading material preserves legal privilege.
Confidential matters are restricted to designated people, with those restrictions applied to files in Xakia's integrated repository or connected document system. Roles, teams, locations and individual permissions further limit access, while outside counsel receive only material shared through Connect. The service agreement protects subscriber information, and the AI addendum prohibits training on inputs and outputs by Xakia or its model providers. A separate clause permits training on anonymous data and statistics derived from inputs. Inputs are processed transiently, with no Xakia retention after delivery. These controls address matter confidentiality and privilege, but unnamed third party models and their specific handling remain unresolved.
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.
Output is not legal or other professional advice under LawVu's AI service terms, which require review before decisions or further processing. The acceptable use policy prohibits substituting the service for licensed advice, allowing unqualified people to rely on legal outputs without professional supervision, or making legal determinations solely from AI. The Business Portal preview turns conversations into legal requests when human input is needed and identifies the customer's legal team as the source of legal advice. It currently serves in-house users acting as proxies, with standard business users excluded from that beta. These are concrete professional boundaries within an internal legal workflow. They do not vary by jurisdiction or map individual features to local rules on authorized legal practice.
AI output from Xakia is general information rather than legal advice. Its AI addendum requires use by legal practitioners or under their direct supervision, in accordance with professional conduct obligations. Users must validate or disregard output using their own skill, judgment and experience, and a lawyer must review it as appropriate to the intended use. Automated decisions with legal or similarly significant effects require human review. These obligations address professional supervision alongside the product's drafting and analysis functions. They do not set jurisdiction specific limits on what legal work the models can perform or map the controls to particular bar guidance.
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.
Six AI principles at LawVu address user needs, transparency, reliability, security, trust and privacy. They apply to AI in the business and its products and are intended to be reviewed annually. Its AI FAQ commits to privacy impact assessments for new features and places AI within the existing security and privacy program. Organization administrators control generative AI enablement; some non-generative functions, including invoice extraction, are enabled by default. Workflow administrators can test actions and inspect execution outcomes before broader use. The MCP guide explicitly describes prompt injection, tool impersonation and external-provider risks. Neither those guides nor the principles provide bias test results or a named AI governance owner. ISO 42001 certification claims also conflict with the older FAQ's statement that an audit is still being pursued.
AI controls at Xakia sit in the features and the contract rather than in a governance framework. Its AI page makes every feature opt in, so the customer decides what its team uses and when, and gives each feature a human checkpoint so nothing is applied without a user's sign off. The AI addendum bars automated decisions with legal or similarly significant effects on individuals without proper human review, and bars uses that are discriminatory or bias inducing. It also requires a lawyer to review output. The feature set includes reviewed contract revisions, extracted terms and invoice findings, but no evaluation method accompanies those workflows. Xakia sets out no AI principles, owner of AI governance, testing before release or bias findings, holds no ISO/IEC 42001 certification, and its security page covers information security.
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.
The data processing addendum requires breach notice without undue delay and, where feasible, within 72 hours. New subprocessors require thirty days' advance notice, with fifteen days to object on data protection grounds. Personal information must be returned or destroyed at termination or request, subject to the agreement and legal retention. The SaaS terms allow a request within one month after termination and backup retention for up to one year. Azure inference is transient, but Assistant conversations, actions and approvals persist: logs are visible for thirty days and available from LawVu afterward, without a stated final deletion date. Restricted membership, SAML authentication with SCIM provisioning, enforceable multifactor authentication, AES-256 encryption and TLS 1.2 or later control access. Backups provide a 35-day point-in-time restore.
Subscriber data stays for thirty days after termination before deletion under Xakia's agreement, while the AI addendum treats input and output processing as transient and leaves necessary retention to the customer. Backups run every five minutes and remain for thirty five days in the same country. The DPA requires breach notice without undue delay and thirty days' advance notice of new subprocessors, with objection rights. Encryption, restricted matters, multifactor authentication and annual penetration tests support access protection. The processor schedule identifies hosting and business systems but does not name the AI model providers. The service terms also exclude responsibility for sensitive data, including protected health information.
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 SaaS terms, which apply to order forms signed after 13 May 2024, cap each party's liability in a calendar year at the fees paid or payable for that year. They exclude indirect and consequential loss, lost profits and loss of data, and lift the cap only for personal injury, death or fraud. LawVu indemnifies the customer against third party claims that its use of the services or platform infringes intellectual property. The indemnity excludes unauthorized use or modification, breach of the agreement, materials LawVu did not supply, and third party or personal data. The AI service terms, effective 30 April 2026, carve AI output out. LawVu does not warrant that outputs are complete or accurate, accepts no liability for loss relating to them, and requires the customer to review every output before use. The services are provided as is. New Zealand law governs, or Delaware law for customers in the United States, with arbitration in each.
Warranty disclaimers and liability limits apply to the AI features through Xakia's platform agreement and AI addendum. Each party's liability is capped at the amounts paid or payable in the 12 months before the first incident. Xakia's liability for breaching its data security or confidentiality obligations, or any data processing addendum, rises to three times that cap. Indirect, consequential and lost profit damages are excluded. Its intellectual property indemnity is outside the general cap. Xakia defends the subscriber and its users against third party claims that the platform, used as authorized, infringes intellectual property. Trials and betas carry no warranty or indemnity, and Xakia's liability for them is capped at $1,000. The AI addendum says output may not be accurate or up to date and must be reviewed by a lawyer, and Xakia gives no indemnity specific to AI output.
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.
MCP setup in LawVu requires support activation, followed by an organization administrator enabling the integration tile. External AI clients connect to its remote endpoint with OAuth 2.0. Read tools retrieve matters, contract details, files and knowledge; write tools create matters or wizard contracts, assign tasks, post status updates and trigger contract workflows. Each action respects record membership and enters the activity log. Microsoft's Power Platform connector provides a separate commercial authentication route and distinct United States Government and demo connections; connections are not shared between users. Draft operates inside Word and supports side-by-side Word or PDF comparison. Sizzling Platter's customer account describes linking matters to SharePoint folders and using the API for Power BI. The buyer controls the security of external AI clients and licenses for third-party platforms.
Integrations are listed by plan on the pricing page. Every plan includes Outlook, Word, SharePoint, Gmail and Microsoft Teams integrations, API access, and single sign on through Entra, Okta, SCIM or ADFS. Professional and above add iManage and NetDocuments integrations and connections to Brightflag, DocuSign, Ironclad and Josef. Slack and Salesforce appear on the higher plans, with Salesforce at extra cost. Xakia offers to keep documents in the customer's own SharePoint rather than migrating them, and the API carries no charge. The pages name each connection without saying what data moves in which direction or what an administrator configures. An MCP server with connectors to Claude and Copilot is listed as coming soon for All In.
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.
Customer data sits in Microsoft Azure, and LawVu contractually commits to the chosen storage region: Australia, the United States, Europe, Canada or the UAE. Primary and secondary data centers remain within that geographic boundary, with logical segregation between customers. The subprocessor list specifies a separate route for some AI processing. United States and European Azure OpenAI data stacks follow the SaaS agreement; Australian and UAE customers' requests run in Europe, and Canadian requests run in the United States. The AI FAQ describes that processing as transient, while the customer's system of records stays in its selected region. A customer-selected AI provider connected through MCP receives relevant data under its own terms, outside that native inference arrangement. The service availability commitment is 99.95 percent.
Microsoft Azure hosts Xakia, and customers choose where data is stored in Australia, Canada, the Netherlands, the United Kingdom or the United States, with a Middle East location listed as coming soon. Different users in one account can have different data locations. European and UK customers can keep data on Azure servers in the United Kingdom or the Netherlands, and the Terms of Service, effective 1 October 2025, are built around the data location the subscriber selects. Section 1.4 of the AI addendum says AI processing runs where the third party AI tools are available and may differ from the selected location, with the processing location shown in the subscriber's account. Separate privacy policies cover Australia, the United States, Europe and Canada. Whether customers share infrastructure is not stated.
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.
LawVu dates its ISO 27001 certification to April 2019, its first SOC 2 audit to January 2020 and its first SOC 1 audit to February 2021. It also claims 2025 certification against Dubai's cloud service provider standard and AssuranceLab compliance seals for HIPAA, GDPR and CCPA. The certificates, statement of applicability, SOC reports and penetration test reports require access to a confidential security pack. The open pages do not identify the current ISO or SOC auditor and reporting period. Annual penetration testing uses CREST-certified providers, with daily Qualys vulnerability scans. AI marketing now claims ISO 42001 certification, while the March 2026 AI FAQ still describes certification as being pursued. That unresolved difference prevents treating the newer claim as a verified certificate.
Xakia states certification to ISO/IEC 27001:2022, SOC 2 Type 2 and HIPAA compliance on its security page, FAQ and pricing page, with regular independent audits of its applications, systems and networks. Its trust center shows compliance monitoring and controls and offers the ISO 27001:2022 certificate, a penetration test attestation and the SOC 2 Type 2 report on request. The data processing agreement and current and past subprocessor lists sit beside them. Third party experts test the application and infrastructure once a year, and third party tools scan the platform for vulnerabilities continuously. The platform is hosted on Microsoft Azure in Tier III facilities certified to ISO 27001:2022, SOC 2, PCI DSS, HIPAA and FedRAMP, with Australia's IRAP and the UK's G Cloud among the regional standards. The open pages name no auditing firm, audit period or report date.
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.
Microsoft services are named feature by feature. Azure AI Foundry supplies models for contract extraction, bulk imports, billing guidelines and drafting. Azure Document Intelligence handles invoice extraction; Content Understanding also supports Workspace Intelligence and billing guidelines, and Azure's retrieval service supports Draft. The Assistant uses Microsoft's Agentic Framework with Azure OpenAI, while Lens combines Azure Document Intelligence and Azure OpenAI models. LawVu gives no specific model versions for those features and no advance notice of model changes. LawVu provisions and manages the native models. Its separate MCP server connects customer-chosen tools such as Claude, ChatGPT and Copilot agents, with provider processing outside LawVu's control. The subprocessor list maps Azure OpenAI processing countries, and the DPA adds notice before new subprocessors.
Third party AI tools, defined as the third party language models and AI technology behind the AI features, appear in Xakia's AI addendum without being named. It says Xakia does not control their operation or continued availability, and that their processing location may differ from the subscriber's data location and is shown in the subscriber's account. Section 6 says any flow down terms the providers require are displayed at trust.xakiatech.com, and section 8.3 gives at least thirty days' notice of changes to the addendum. The subprocessor list in Xakia's trust center names Microsoft and business systems such as HubSpot and Stripe but no AI model provider. Neither the third party AI tools nor Xakia may train models on customer inputs or outputs.
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.
Subscriptions are quoted from selected capabilities and user numbers, with no starting amount or standard seat rate. Matters covers intake and matter management, Essentials adds contracts, and Ultimate adds spend. Draft can be purchased separately or added to a subscription; the agentic workflow builder is a paid addition. The plans FAQ specifies a first-year onboarding fee, while support and customer success are included. Some integrations cost extra. Spend pricing uses the number of outside firms rather than a percentage of legal expenditure. The invitation AI beta is initially available to selected Essentials and Ultimate customers at no additional cost, with future consumption tiers under evaluation. Standard terms renew annually unless either party gives thirty days' notice; renewal increases generally have a fifteen percent ceiling, excluding consumption-based rate cards.
The interactive pricing schedule lists Advance at $100, Professional at $125, Enterprise at $155 and All In at $230 per user monthly, with USD as the default currency and alternatives for other regions. Advance adds $50 per user monthly for AI and Professional adds $30, with usage caps. Enterprise includes capped AI and All In removes caps subject to fair use. The linked machine readable schedule still gives lower base prices and older availability descriptions. Onboarding has a separately quoted fee. Monthly service periods require thirty days' nonrenewal notice under the agreement, while annual billing offers a discount. Outside counsel use Connect without a portal or invoice percentage charge.
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.
LawVu serves in-house legal departments across healthcare, higher education, software, financial services, government, energy, manufacturing and insurance. Its role-specific workflows address general counsel, legal operations, IT and teams adopting their first legal system. Customer groupings span Asia-Pacific, North America, the United Kingdom, Europe and the UAE. LegalOS combines intake, matters, contracts, files and reporting, with outside counsel spend in the Ultimate plan. Law firms participate as service providers through matters, invoice submission and billing guidelines; this is not a law firm's client-accounting or court-filing system. Draft supports common commercial agreements such as NDAs, MSAs, supplier and sales contracts inside Word. The Business Portal AI remains a restricted preview, so its broad intended business audience should not be mistaken for general availability.
In house legal teams of roughly 2 to 200 people are the target, a range Xakia repeats on its pricing page. The platform can be configured in English, Japanese, Spanish, French, Danish, Dutch and Brazilian Portuguese, and the spend module supports all currencies. Xakia has offices in Melbourne, Kansas City and London and says it has many hundreds of clients worldwide. Scope is stated by module, covering matters, intake, contracts, spend, documents, entities, disputes and reporting, with a stakeholder license for non legal users. Customers store data in Australia, Canada, the Netherlands, the United Kingdom or the United States and can set different data locations for different users. Industry segments and practice areas are not described.
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?
Under the AI service terms, part of the customer agreement, third parties may not use customer data or outputs to improve or train AI and machine learning models, including LawVu's AI services. The AI security and privacy FAQ adds that LawVu trains no models on customer data, directly or indirectly, and does not let its service providers train theirs. Inputs and outputs to the AI services are not used for training. LawVu operates under an exemption from Microsoft's standard Azure OpenAI monitoring, so Microsoft may not access or monitor that usage.
The SaaS terms let LawVu collect anonymized statistical data from use of the services, such as session data, page loads and which features are used, to improve services and develop new ones. This native-service commitment does not reach a customer's separately connected MCP provider, whose own terms apply.
Xakia's AI addendum bars training on customer content and permits training on anonymized data derived from it. The addendum, effective 1 February 2026, forms part of the Terms of Service. Section 2.4 says neither Xakia nor its third party AI tools will train any AI model on inputs or outputs, and that using them to improve service features is not training. Section 2.5 lets Xakia collect derived data, defined as anonymous data and statistics derived from inputs.
Xakia may use that data for benchmarking, reports and training its AI features or AI models, and discloses it outside Xakia only in anonymized, aggregated form. Section 1.3 lets a customer opt out of the AI features in its account, and the addendum offers no separate opt out from derived data. Xakia does not name the third party AI tools behind the features, and their processing may run outside the customer's selected data location, as shown in the customer's account.
Prompt and Output Retention
How long does the product keep what a lawyer typed, and can that be set to zero?
Native AI requests are transient at the inference service, while the results and conversations can persist inside LawVu. Assistant conversations, actions and approvals form an audit record visible for thirty days, with older logs available from LawVu on request and no final retention period specified. Business Portal conversations remain available to the user and authorized administrators. Changing a user's matter permissions does not revise prior conversation history.
The SaaS terms allow a copy request within one month after termination and backup retention for up to a year. External AI tools connected through MCP have separate retention terms controlled by the customer's chosen provider.
Section 2.8 of Xakia's AI addendum sets the retention period for AI prompts and outputs at zero. Because the AI features process data ephemerally, Xakia does not retain inputs or outputs after the output is delivered. Inputs are processed transiently and not stored, and the customer is responsible for capturing anything it needs to keep. Xakia fixes the period rather than letting the customer configure it. It governs the AI features rather than matter data held in the platform, which the Terms of Service keep for 30 days after termination before permanent deletion.
Extracted key terms enter the contract repository only when a user clicks Apply, and then follow the matter data's retention.
Ethical Walls and Matter Segregation
Does retrieval respect the firm’s ethical walls, or can the model read across them?
A restricted matter admits only named members, including external legal service providers where invited. Administrators can see limited identifying fields in a restricted-entities report, but not the matter contents. Contracts, statements of work and amendments created within restricted parents inherit membership; previously linked contracts and newly created child matters do not automatically inherit the restriction.
Delegates also require explicit membership. The native Assistant and MCP connections apply the user's access rights, using LawVu's own permission model. Saved conversations are not revised after later permission changes, so a user can retain previously discussed material in chat history.
Role based permissions control access by user, team or matter, Xakia says sensitive matters can be restricted to specific people, and outside counsel see only what the customer chooses to share through the Connect portal. Single sign on through the customer's own identity provider handles authentication. The product runs its own permission structure rather than applying a document management system's access model at query time, and Xakia does not say whether the AI features apply those permissions when they read documents.
Third Party Request and Subpoena Notice
If someone subpoenas the vendor for a firm’s data, does the firm hear about it first?
Any legally binding request by a regulatory body for disclosure of customer personal information triggers prompt notice to the customer under LawVu's data processing addendum, unless the law prohibits notice. The commitment covers requests addressed to LawVu or any of its subprocessors. The addendum also commits LawVu to tell the customer about monitoring or investigations by a regulatory body. Regulatory bodies are defined to include government departments and regulatory and statutory bodies entitled to regulate or investigate data security and privacy.
The SaaS terms add that a party required by law to disclose the other's confidential information must notify the other immediately in writing where permissible and disclose only what is required. LawVu issues no transparency report. Once a calendar year, under a nondisclosure agreement, a customer can have the latest independent audit report, with an audit at its own cost where the report is not enough.
Any third party request made directly to Xakia about its processing of the subscriber's personal data must be reported to the subscriber promptly, with details of the request, under the data processing addendum. The commitment applies to the extent the law permits. Third party requests include those from a regulatory authority. Xakia will not respond to such a request without prior notice to the subscriber and a chance to object, except where the law requires it or to confirm that the request relates to that subscriber. The commitment covers personal data under the addendum, and Xakia issues no transparency report.
Primary Law Corpus Provenance
Where does the law in this product come from, and does the vendor have the right to use it?
LawVu's AI works from the customer's own material. The Assistant draws on matters, contracts, policies, knowledge repositories and legal notes in the customer's LegalOS. LawVu Draft drafts and reviews from the customer's clause library, precedents and imported playbooks, with LawVu's base templates as a starting point. The Business Portal assistant answers business users from knowledge content the legal team makes available, and LawVu Lens reads the customer's executed contracts. No primary law collection, licensed legal content or update cadence sits behind the answers.
Xakia holds no legal content collection. Its AI features work on the customer's own contracts, documents and invoices in the platform, so the material and its provenance are the customer's. Contract review applies the customer's own playbook and also checks a contract against governing law, market practice and basic drafting standards, without naming where that knowledge comes from. Xakia supplies no licensed legal content and states no license basis or update cadence for the knowledge its model applies.
Good Law Verification
Does the product tell you when the authority it just cited has been overruled?
LawVu works on contracts, matters and invoices rather than case law, so a citator sits outside the product. Draft checks contracts against the customer's playbooks and clause library, and Lens cites back to the source contract rather than to legal authority. Lens identifies risky terms and inconsistencies across a contract portfolio against the customer's own standards. Keeping playbooks, clauses and policies in step with changes in law is the work of the legal team that maintains them.
An in house team's matters, contracts and spend, not case law, are what Xakia manages, so a citator sits outside the product. Its AI reviews and summarizes the customer's own contracts and extracts key terms, and its query tools work across the customer's own files. Keeping contract standards in step with changes in law is the legal team's work, and the AI addendum makes users responsible for checking that output is accurate and up to date.
Refusal and Uncertainty Behavior
What does the product do when the answer is not in the corpus?
Lens distinguishes a provision it cannot find from language it cannot confidently interpret. Not Found means no relevant clause was identified; Unclear marks ambiguous or conflicting wording that cannot be mapped to a defined value. Reviewers can open the source contract and inspect the text, with citation highlighting for PDFs. The Business Portal preview also declines unsupported answers, suggests another retrieval attempt or escalates to a legal request.
These are stated product behaviors, not an independently measured refusal rate. The AI terms still require review, and the current preview is available only to in-house proxy users.
Each value from key terms extraction comes with a confidence indication, the clause it came from and the exact text it relied on. A user reviews and corrects each value before clicking Apply, so nothing reaches the contract repository unchecked. Contract review flags deviations from the playbook for a lawyer to decide, and the generated query remains visible for correction and reuse. Xakia describes no path for declining a request when a document holds no answer, such as a key term the contract does not contain.
The AI addendum tells users to validate or ignore output using their own professional judgment, because it may not be accurate or up to date.
Fabricated Citation Record
Does a public court record exist addressing fabricated or hallucinated legal citations in output from this product?
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 LawVu, LawVu Draft, LawVu Lens or ClauseBase. LawVu drafts contracts and answers questions from an in house team's own records rather than producing court filings.
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 Xakia. Xakia's AI produces contract redlines, summaries and extracted terms from the customer's own documents rather than court filings.
Bar Guidance Alignment
Has the vendor engaged in public with the ethics opinions its buyers are bound by?
Professional duties appear in LawVu's acceptable use policy in general terms. It bars using the AI services as a substitute for licensed legal, financial or other professional advice. It also bars letting unqualified individuals generate or rely on outputs such as legal documents without appropriate professional legal supervision, and relying solely on AI output for legal or regulatory determinations without human review.
The policy incorporates Microsoft's Enterprise AI Services Code of Conduct and forms part of the SaaS agreement. LawVu's guide to AI for in house teams covers adoption and procurement. These terms do not name a bar opinion, such as ABA Formal Opinion 512, or a professional conduct rule in New Zealand, Australia or elsewhere.
Professional duties appear in the AI addendum in general terms. It requires the AI features to be used only by or under the direct supervision of legal practitioners in accordance with applicable professional conduct rules. It also bars automated decisions with legal or similarly significant effects on individuals without proper human review in compliance with applicable professional ethics, guidelines and rules. Neither the addendum nor Xakia's site names a bar opinion, such as ABA Formal Opinion 512, or guidance from the Law Council of Australia or a state law society, in Xakia's home market. Xakia's in house hub covers evaluating matter management software and legal operations practice.
Billing and Fee Posture
Does the vendor address what happens to the bill when the work takes an hour instead of six?
Outside counsel fees are recorded on the payer's side, in LawVu's spend module. Firms upload invoices, including LEDES files, and see their matters, invoices and billing guidelines in a self service portal. The legal team receives, adjusts and approves invoices, routes approvals through drag and drop workflows and sends approved invoices to finance. AI extracts cost details from PDF invoices and flags discrepancies with billing guidelines, such as rate or scope issues, in real time.
Reporting tracks spend by firm and matter. Legal teams send RFPs with scope and budget, compare firm proposals side by side and receive AI provider recommendations. Spend management is priced by the number of firms a customer works with rather than by its spend. The billing guidelines are the customer's own, so any rule on how firms bill AI assisted work is set by the legal team in those guidelines.
Xakia's spend module is the payer's record of outside counsel fees. Outside firms accept work, update matter details and submit invoices through the Connect portal, which they use at no charge, and the legal team tracks spend against matter budgets in any currency. Xakia sets no limits on the number of firms, invoices or spend managed. It takes no percentage of invoices, which it calls a clip of the ticket, so its revenue does not grow with a customer's legal spend.
Budget tracking compares forecasts with spend to reduce overspend. AI invoice review now checks billing rules and discrepancies in totals, tax and vendor details; reviewers waive or act on findings. How firms bill AI assisted work is left to the customer's own billing terms.
Outside Counsel Guideline Readiness
Can a firm get this vendor through a client’s AI clause without a bespoke negotiation?
LawVu supplies a provider list with processing countries, an AI feature-to-provider mapping, a DPA and AI service terms. The subprocessor list distinguishes mandatory infrastructure from optional services and includes Azure OpenAI routing for each customer storage region. Its DPA adds incident notice, subprocessor objection and annual independent audit-report access. Certificates and detailed reports require the confidential security pack.
The MCP guide separately explains that connected AI providers receive customer data under their own terms. This gives a legal team material for a client or auditor disclosure, but the team must add the particular external AI services it enables and resolve the conflicting ISO 42001 claims.
A processor schedule identifies the country and purpose of providers including ChurnZero, HubSpot, Intercom, Microsoft, Stripe, Twilio and Xero. The DPA, AI addendum and security controls are available for diligence, while the ISO certificate, penetration test attestation and SOC 2 report require a request. No AI model provider is identified in the processor schedule despite third party model use under the addendum. Matter restrictions and the limited Connect view control disclosure to external firms, but do not by themselves establish compliance with every client's AI requirements.
Court Disclosure Support
If a judge’s standing order requires an AI disclosure, can the product produce one?
Assistant conversations, approvals and actions enter LawVu's audit logs, as do writes from connected MCP tools. Logs remain visible in the product for thirty days and available from LawVu afterward on request. Lens can export extracted clauses and values to Excel, respecting each user's contract permissions; a reviewer can open the original contract to verify an extraction. These records help explain an internal legal workflow, but do not form a court-specific declaration identifying the model and human review for every submitted document. LawVu primarily manages contracts, matters and invoices rather than filing generated legal arguments.
Parts of the record a court or client might ask for are kept. Extracted key terms carry the clause they came from and the exact text relied on, and reach the contract repository only after a user reviews them and clicks Apply. Xakia's AI page says each decision a user takes on an invoice review finding lands in the audit trail. The platform also records matter activity, intake routing and spend against each matter. No export ties an AI output to the model used, and no model is named.
Xakia's output is internal records, contracts and reporting rather than court filings, so a judicial order on AI disclosure is less likely to reach it.
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.
- Primary Law Corpus Provenance
- Good Law Verification
Which one fits
Choose LawVu if
- Your legal team lives inside Microsoft. LawVu Draft drafts and redlines inside Word from the customer's clauses, precedents and playbooks. Microsoft's connector reference lists LawVu for Power Automate, Power Apps, Copilot Studio and Logic Apps, with 26 actions and 14 triggers. An MCP server lets Claude, ChatGPT and Microsoft Copilot read and write LawVu data within each user's access, and every action lands in the activity log.
- You want assistants that act inside the system. LawVu's Assistant answers questions and creates or updates records once a user approves, and every Assistant action lands in an audit trail. The Business Portal assistant answers business users from knowledge the legal team makes available and recommends a legal request when it cannot find a reliable answer. An agentic workflow builder, sold as an add on and marked coming soon in the help center, runs workflows inside guardrails the legal team sets.
- You manage outside counsel spend at scale. LawVu's spend module takes LEDES and PDF invoices through a self service portal, extracts cost details with AI and flags discrepancies with the customer's billing guidelines. Legal teams send RFPs and compare firm proposals side by side, and spend management is priced by the number of firms a customer works with. LawVu names Discord, Etsy, Employment Hero, The Pokémon Company and Estée Lauder among its customers.
Choose Xakia if
- You want to know the price before the call. Xakia lists Advance at $100, Professional at $125, Enterprise at $155 and All In at $230 per user a month, month to month or annual with up to 10 percent off. AI adds $50 per user a month on Advance and $30 on Professional, and a 14 day trial needs no payment card. Onboarding is a separate fee quoted on request.
- Some matters cannot be seen by the whole team. Xakia's role based permissions control access at user, team or matter level, sensitive matters can be restricted to named individuals, and outside counsel see only what the customer shares through Xakia Connect. Different users in one account can keep data in different locations, across Australia, Canada, the Netherlands, the United Kingdom and the United States.
- You do not want your vendor's fee to scale with your legal spend. Xakia charges outside counsel nothing to use the Connect portal, sets no limit on invoices received and takes no percentage of invoice value, which it calls a clip of the ticket. Internal requesters carry no charge, and a stakeholder license gives non legal users access to contracts and entities at a flat annual rate.
In summary
LawVu
LawVu is a legal operating system for in house legal teams from Tauranga, New Zealand, marketed as LegalOS since June 2026. Three plans combine intake with AI triage, matter management, contract lifecycle management and spend management with electronic billing. LawVu Intelligence adds an Assistant for the legal team, a legal assistant for business users, contract and invoice extraction and an agentic workflow builder sold as an add on. LawVu Draft, formerly ClauseBase, drafts and redlines in Word, and LawVu Lens analyzes executed contracts with citations to the source. AI runs on Microsoft Azure AI Foundry and Azure OpenAI, and customers choose a data region in Australia, the United States, Europe, Canada or the United Arab Emirates. LawVu publishes its SaaS terms, AI service terms, data processing addendum and subprocessor list.
Xakia
Xakia is a matter management and legal operations platform for in house teams of roughly 2 to 200 people, founded in 2016 in Melbourne, with offices in Kansas City and London. Modules cover matters and tasks, intake and triage, contracts, spend and budgets, documents, entities, disputes and reporting, and outside firms work through the Xakia Connect portal at no charge. Xakia AI reviews and redlines contracts in Word against the customer's playbook, summarizes contracts and extracts key terms with a confidence score and source clause. Smart Search turns a plain English request into a query the user can edit. Plans run from $100 to $230 per user a month, month to month, with AI an add on for the lower two. Customers store data on Microsoft Azure in Australia, Canada, the Netherlands, the United Kingdom or the United States. Xakia's AI addendum states that output is not legal advice and must be used under a lawyer's supervision, and keeps no inputs or outputs once output is delivered. Xakia holds ISO 27001:2022 and SOC 2 Type 2 certification, and publishes its Terms of Service, data processing addendum and subprocessor list.
Questions buyers ask
LawVu vs Xakia: which is better for a small in house legal team?
Xakia is built for teams of 2 to 200, lists per user prices from $100 a month, runs month to month and sells AI as an add on below its Enterprise plan. LawVu suits a team that wants deep Microsoft integration, assistants that act on records with a user's approval, spend management priced by firm count and a choice of five data regions. Both publish their customer terms and AI terms. From the AI Legal Index, based on each vendor's own published materials as of October 9, 2026. No vendor pays for placement.
How is Xakia priced?
Xakia lists four plans per user per month, with Advance at $100, Professional at $125, Enterprise at $155 and All In at $230. A currency selector shows prices in Australian dollars, United States dollars, euros or pounds. Plans run month to month with no lock in, and annual billing takes up to 10 percent off. AI adds $50 per user a month on Advance and $30 on Professional, and is included with usage caps on Enterprise and without caps on All In. Onboarding is a separate one off fee, and a 14 day trial needs no payment card. From the AI Legal Index, based on each vendor's own published materials as of October 9, 2026. No vendor pays for placement.
What security documentation can you get from LawVu?
LawVu's security pack holds the ISO 27001 statement of applicability and certificates, SOC reports, penetration test reports and the business continuity plan. It is released through a request form after a notice of confidentiality, or by email under a nondisclosure agreement. LawVu has held ISO 27001 since April 2019 and dates its first SOC 2 audit to January 2020 and its first SOC 1 audit to February 2021. AssuranceLab, which issued its HIPAA, CCPA and GDPR seals, is the only assessor named on its open pages. From the AI Legal Index, based on each vendor's own published materials as of October 9, 2026. No vendor pays for placement.
Can you use either platform without the AI?
Yes, both. Xakia sells AI as an add on for Advance and Professional and makes every AI feature opt in, so a team can run matters, intake, contracts and spend without it. LawVu's generative AI features are switched on by opting in, and administrators enable or disable new features at any time. LawVu relaunched the platform as LegalOS around its AI in June 2026, while its intake, matters, contracts and spend modules run as workflow without a model. From the AI Legal Index, based on each vendor's own published materials as of October 9, 2026. No vendor pays for placement.
What do LawVu and Xakia both leave unpublished?
Neither publishes an accuracy rate or test results for its AI, and neither offers an indemnity covering AI output, since LawVu's AI terms exclude liability for output and Xakia's addendum applies its general liability limits. Neither names a bar opinion on AI or issues a transparency report on government requests. Neither names the model version behind most of its AI features, and Xakia names no model provider at all, while LawVu names Azure OpenAI. From the AI Legal Index, based on each vendor's own published materials as of October 9, 2026. No vendor pays for placement.
Neither vendor paid for inclusion, placement or a grade, and neither reviewed this page before it published. Everything on it comes from public material on the dates shown. How the index grades.
LawVu's AI product pages say it is certified to ISO 42001, while its AI security and privacy FAQ, updated March 2026, says it is aligned with the standard and pursuing certification. LawVu's AI service terms bar third parties from training on customer data or outputs, including LawVu's own AI services. Xakia's AI addendum bars training on inputs and outputs but lets Xakia train its AI features on anonymized data derived from them. Opting out of the AI features is the only way to avoid that use of customer data. Xakia's AI page describes invoice review as live, while its plan comparison lists AI invoice review as coming soon. Neither vendor reviewed this page.