Dazychain vs Xakia: how they compare in 2026
Dazychain and Xakia are both matter management platforms for in house legal teams, both from Melbourne, and both publish per user prices. Dazychain is made by Yarris Technologies. Dazychain sits in the top two bands on ten of fifteen axes and Xakia on six of fifteen, identical on nine. Dazychain's lead is how tightly its AI is bounded. Its AI runs only when a user clicks and reads only the matter or document it was opened on. It can be switched off for an account and runs on Amazon Bedrock in Australia or the United States, with nothing stored. Dazychain publishes no customer agreement at all. Xakia's counterweight is an AI addendum in its contract. It states that AI output is not legal advice, must be used under a qualified lawyer's supervision, and is not retained once delivered. The same addendum bars training on inputs and outputs but permits training on anonymized data derived from them.
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.
Artificial intelligence is present, useful and peripheral, and the vendor's own design decisions say so. The product is sold first as matter management, described in its own footer strapline as capture, triage, manage, resolve, and the pricing table shows a complete platform at every tier with AI as one section among matter management, documents, spend, collaboration, workflow, reporting and integrations. Three published facts place it here. The AI is capped by usage on the entry tier and unlimited above it, so it is priced as an add-on rather than as the mechanism. It runs only when a user engages a function by clicking a button, so nothing happens by default. And the FAQ states that AI can be completely turned off for an account, leaving a working matter management system behind. What the AI does is summarise, search, answer questions and compare a contract to a playbook, all of which sit on top of the record rather than constituting it. Checked 4 September 2026.
FIRST C ON THIS AXIS IN THE INDEX. Artificial intelligence is present but peripheral: a feature layer on a product whose value stands without it. The decisive evidence is commercial rather than interpretive. The vendor sells its AI capabilities as a priced add on for the Advance and Professional tiers, included only in the Enterprise and All-In tiers, so a substantial share of its customers run the product with the AI switched off entirely and the vendor prices on that basis. That is the clearest possible statement that the core value stands alone. What the core is: matter management across the lifecycle, intake and triage, contracts, spend and budget management, document management, entity management and reporting, all of which are workflow and record keeping capabilities operating with no model behind them. MEMBERSHIP: the AI bar is nonetheless cleared, and this record is properly enrolled. Shipped AI features exist and are named individually: contract review and redlining, contract summarisation, key terms extraction, smart search and invoice review. The vendor's own framing is candid, describing AI as something it is continuing to expand rather than as what the product is.
Citation Accuracy and Hallucination Disclosure
Whether the vendor publishes measured accuracy on citations and assertions, grounds output to primary sources, and says plainly what its system does when it does not know. Legal has a documented public record of fabricated citations reaching filed briefs, so an untested claim of accuracy is not evidence.
Accuracy is asserted with an unusually candid caveat and nothing is measured. The FAQ states that AI extractions are designed to be highly accurate but, like any automated system, may require review, and that users can validate and edit extractions directly in the platform before anything is finalised or shared. A limitations entry adds that the AI works only within the matter or document it is engaged on and that clear, well-structured and complete documents lead to better results, which is a real statement of input sensitivity that most vendors omit. Grounding scope is therefore described. What is absent is measurement of any kind: no accuracy figure, no test set, no evaluation, no error rate, and no description of how an extraction traces to the passage it came from. The one hallucination reference is not the vendor's own: the security section attributes minimisation of model hallucinations to Amazon Bedrock, which is a claim about the infrastructure rather than a published position on the product's output.
Accuracy is asserted without measurement and without a described grounding method. The vendor's published position is that it expands AI capabilities with a focus on things that genuinely save legal teams time without compromising accuracy or security, which asserts accuracy as a constraint on its roadmap rather than reporting it as a result. Searched the platform pages, the FAQ, the security page, the pricing page and the in house hub articles on 29 Aug 2026 and located no accuracy figure, no hallucination rate, no test set, no evaluation methodology and no independent benchmark participation. Nothing describes how contract review, redlining, summarisation or key terms extraction ground their output, whether extracted terms link back to the clause they came from, or what a reviewer sees to check a suggestion. For a product performing redlining and key term extraction on contracts the customer will sign, the absence of any described verification surface is the material gap.
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 system does nothing on its own, and that is published as a design commitment rather than implied. The FAQ states that AI accesses data only when a user actively engages an AI function such as clicking a button, so privileged and confidential matters remain untouched until assistance is explicitly requested; that the AI works only within the matter or document it is engaged on and does not reach data outside that context; and that extractions may require review, with users validating and editing them in the platform before anything is finalised or shared. AI can also be turned off entirely for an account. That is a genuine control structure with a real review point. What holds it below the top band is that the constraints are binary rather than graduated: no confidence signal is surfaced against an individual extraction, no threshold or abstention state is described, and the FAQ question asking whether AI can be restricted by role or matter type is answered only as to switching it off for the whole account, leaving the finer-grained half of its own question unanswered.
Autonomy is limited by design and the control structure is implied by workflow rather than published. The product's shape places the human at every decisive point: legal teams route and tag intake requests themselves, set up matters from templates, assign tasks, and the AI features act on documents the user is working in, offering redlines, summaries and extracted terms rather than executing anything. Nothing in the located material describes an agent that acts unattended. But that is a description of a workflow rather than a published oversight position. Searched the platform pages, the FAQ, the security page and the in house hub on 29 Aug 2026 and located no statement of what the AI features decide unaided, no review surface described as such, no threshold at which anything escalates, and no statement of what the vendor commits to when an output is wrong.
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.
Named customers with named people in stated roles, and no figures. Two named legal officers are quoted with linked case studies: Helen Carr, Deputy General Counsel at MYOB, describing using Dazychain's matter data to demonstrate an increase in the volume and complexity of matters and to advocate successfully for additional headcount, and Mercia Chapman, Senior Legal Counsel at Equity Trustees, on email capture and retrieval. A Client Stories section carries further interviews and was not opened in this pass. The MYOB account is the stronger of the two because it describes a specific operational outcome that the platform's reporting made possible rather than a satisfaction sentiment. What keeps this at B is measurement: nothing is dated, no quantity is attached to either customer, and the corporate claims run to unquantified language about efficiency. This is materially better deployment evidence than the seed list predicted, which flagged the name as directory-level only.
Testimonials and a logo strip stand in for deployment evidence. Attributed customer quotes are published including one claiming hundreds of hours saved and another that the platform transformed how a team works, alongside a leading brands strip and a stated base of in house legal teams of all sizes worldwide. Independent review presence exists on G2. Searched the site, the in house hub, the pricing page and the review platforms on 29 Aug 2026 and located no named customer paired with figures and a date, no case study with an assessable method, and no adoption count. Hundreds of hours saved is a testimonial rather than a measured result and was not treated as a figure.
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.
Substantive commitments across most of the ground, published on a security page rather than in an agreement. Segregation is stated plainly: data segregation is fully supported and tenant aware, and no user company can access another's information unless permission has been given, as in a collaboration. Encryption is AES-256 at rest in a MongoDB store with HTTPS and TLS 1.3 in transit, documents sit encrypted in Amazon S3, role-based access applies at every tier, and every access and every modification to documents, matters and deliverables is logged and preserved. The AI position is the strongest part and is unusual: the model reaches nothing until a user clicks, it is confined to the matter or document engaged, Amazon Bedrock does not store inputs or outputs, each model runs in a dedicated AWS account, and customer information is not used to train external models. Two things hold it below the top band. No privilege or work product treatment is named anywhere, which the top band requires as its own limb. And there is no customer agreement of any kind on the site, so every commitment here is policy rather than contract.
Substantive published commitments, with segregation unusually well specified, short of the training and retention limbs. Segregation is the strongest element and is published at the level this axis asks for: role based permissions controlling access at individual user, team or matter level, with the vendor stating explicitly that sensitive matters can be restricted to specific individuals and that external parties such as outside counsel see only what the customer shares with them through the Xakia Connect portal. Matter level restriction plus a bounded external party view is a real confidentiality architecture rather than an assertion, and it addresses the in house version of the walls question directly. Single sign on through the customer's existing identity provider is supported. Certification covers ISO 27001, SOC 2 Type 2 and HIPAA with regular independent audits. Two gaps hold this off an A. No statement was located on whether customer content may be used to train or improve models. No retention or deletion terms were located. Attorney client privilege and work product are not addressed directly.
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.
Nothing published addresses the advice line, on a product that drafts documents and answers questions about them. The site inventory was taken from the navigation and footer across four pages on 4 September 2026 and contains no terms of service, no end user agreement and no disclaimer page; the only legal document linked anywhere is a privacy policy. No statement was located that outputs are not legal advice, that a lawyer must review generated content before it is relied on, or that self-service document generation by a business stakeholder carries any limitation. The gap is pointed rather than formal because of one published feature: business clients and non-lawyers complete an intake form which automatically generates a tailored document, such as a non-disclosure agreement, from a pre-approved template and can send it back without manual intervention. That is a document produced for a non-lawyer with no described legal checkpoint, and nothing published sets a boundary around it. No rule of professional conduct, bar authority or jurisdiction limit appears on any surface.
AMENDED 6 September 2026 under R94 on a newly located agreement; the pull-1 C recorded no located advice line. The vendor states what its AI is and is not, who may use it and how it sits with a lawyer's duties, in the agreement itself. Section 5.3 of the Artificial Intelligence Addendum effective 1 February 2026 states that AI output is general information and does not constitute legal advice, that the AI features are not a substitute for legal or professional skill, judgment and experience and must only be used by or under the direct supervision of qualified legal practitioners in accordance with applicable professional conduct rules, and that the user must validate or ignore notifications with their own professional judgment; section 5.4 requires qualified-lawyer review of output; section 3.2(a) bars automated decision-making with legal or significant effects without proper human review in compliance with professional ethics rules. The buyer is an in-house legal team. No specific bar guidance is named, which would have been additional rather than required. AI Addendum read in full 6 September 2026.
AI Governance and Bias Disclosure
Published governance over model behavior: who owns it inside the vendor, what is tested before release, and what is disclosed about disparate output across matter types, parties, or populations.
No governance position was located. There is no responsible AI page, no principles statement, no named owner accountable for model behaviour, no pre-release evaluation regime, no testing results and nothing at all on bias. The AI FAQ is substantial and entirely operational, covering what the AI can do, where it runs, what data it uses, whether it can be turned off and what its limitations are, which is security and scope rather than governance. The security page is detailed and is likewise about information security, which the axis definition treats as a separate subject and which is graded on the stewardship and certification rows rather than counted twice here. The one adjacent artifact is an Information Security Management System with documented policies, annual staff training and documented acceptance, but that governs employee conduct rather than model behaviour. Searched the AI page, the security page, the pricing page, the features index and the site navigation on 4 September 2026.
Searched the site, the FAQ, the security page, the pricing page, the product updates section and the in house hub on 29 Aug 2026. No governance position for model behaviour was located: no AI principles or framework, no named owner of model governance, no pre release testing regime, no AI management certification such as ISO 42001, and nothing on uneven output across matter types, parties or populations. The vendor's only located statement touching governance is that it expands AI capabilities without compromising accuracy or security, which is an intention rather than a mechanism. Recorded as an absence on surfaces that were reached rather than assumed: the security page, FAQ and pricing page were all read and none addresses AI governance. Rebuttable with one link.
AI Safety and Data Stewardship
Retention, deletion, access control, and what happens to prompts and documents after they are processed. Whether the vendor states its subprocessors and its incident practice, or leaves the buyer to assume.
Most of the set is published and specific, with incident practice the missing limb. Retention and deletion are stated together: on account termination customer data is retained for the period the customer requests, usually one month, then deleted, with alternative arrangements available on request. Access control is described at several levels, with role-based access at every pricing tier, an audit trail logging every access and every modification to documents, matters and deliverables, staff police checks on joining and every two years, signed confidentiality agreements for employees, third parties and contractors, and annual security training with documented acceptance. Encryption is AES-256 at rest and TLS 1.3 in transit. Infrastructure suppliers are named rather than gestured at: Amazon for application and file hosting and Object Rocket for the database, both in Australia and the United States. Penetration testing runs internally before each release and annually through an unnamed external agency. What is absent is any breach or incident notification commitment: nothing states whether, when or how a customer would be told, and no subprocessor register or change notice exists.
Certification and access control are published while the rest of the stewardship picture is not. Real and stated: ISO 27001, SOC 2 Type 2 and HIPAA certification, regular comprehensive independent audits of applications, systems and networks, enterprise grade cloud infrastructure, single sign on, role based permissions to matter level, and security documentation available on request with the SOC 2 report obtainable from the team. Searched the security page, the FAQ, the platform pages and the pricing page on 29 Aug 2026 and located no stated retention period or deletion control, no encryption specifics for data at rest or in transit, no named subprocessor list, no hosting provider or region, and no incident or breach notification practice. Recorded at C because access control and certification are covered well and the operational elements this axis names were not located.
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.
Nothing published allocates loss, because no customer agreement is published at all. The site inventory taken from the navigation and footer across four pages on 4 September 2026 lists a privacy policy and a sitemap as the only legal documents; there is no terms of service, no subscription agreement, no end user terms and no service level document anywhere on the site. Consequently no indemnity, no liability cap, no warranty, no exclusion and no insurance position could be located, and a buyer cannot read the allocation of loss before signing even though the product offers a self-serve fourteen-day free trial and publishes per-user prices, which is the commercial posture of a product a buyer might purchase without negotiation. This is a documented absence rather than a retrieval failure: every page fetched rendered in full and the inventory is complete. The one adjacent statement located is a security-page assertion that regular audits keep the cloud environment secure, which is a practice claim rather than a recourse position.
Searched the site, the FAQ, the pricing page, the security page and the in house hub on 29 Aug 2026. No published indemnity, liability cap, carve out, warranty on output or insurance position was located, and no customer terms of service was located as published on the property. Recorded as a pure absence on the surfaces reached. Worth noting the contrast within this record: the vendor is exceptionally clear about commercial terms, publishing tiers, per user pricing, trial conditions, no lock in and explicit statements that outside counsel are never charged and no percentage is taken on invoice value, and says nothing about who bears the loss when an AI redline or extracted term is wrong. Commercial transparency and liability transparency are different things and this record separates them sharply.
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.
Named integrations with tier availability published, and no description of what they move. The pricing comparison table is the best source and lists them by tier: single sign-on through Entra, Okta, SCIM and ADFS, plus DocuSign, Outlook and Gmail at every tier, and API access at the top tier only. That is more useful to a buyer than a logo wall because it answers what a given plan actually includes. Two entries are marked as not yet available and are recorded as such rather than credited: SharePoint integration is listed coming soon at the top tier, and document libraries and knowledge base are coming soon at all three. What is missing for the top band is depth. Nothing states what synchronises, in which direction, on what trigger or what a legal team must configure, and no dedicated legal document management system is live, so a department whose matter files sit in a DMS today has no published path. An integrations page exists and was not opened in this pass.
Integration routes are named without documentation an implementer could use. Published: an open API available free of charge, which the vendor frames as letting customers connect the platform to their existing tools without a commercial gate, and that is a real position since several vendors on this index treat API access as a paid or contact us item. SharePoint is named for document management, with the vendor offering to keep documents in the customer's own SharePoint rather than requiring migration. Single sign on integrates with the customer's existing identity provider. A dedicated LegalTech integrations section is published. Searched those pages and the FAQ on 29 Aug 2026 and located no API documentation reachable without contacting the vendor, no per integration description of what moves in which direction or what an administrator configures, and no other named connector.
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.
Residency is published concretely and tenancy is described only in outline. Locations are given for each layer rather than for the platform as a whole: application data in Amazon data centres in Australia and the United States, the database with Object Rocket in Australia and the United States, and documents in Amazon S3 in the same two regions, with multi-availability zone deployment for redundancy. Processing is addressed separately from storage, which is rare in this lane: the AI runs on Amazon Bedrock infrastructure in Australia or the United States depending on the customer's location, keeping data within the designated geographic boundary, with each model in a dedicated AWS account. Tenancy is described functionally rather than architecturally, with segregation stated to be fully supported and tenant aware, which implies a shared platform with logical separation but never says so. No single-tenant, dedicated or self-hosted option is offered, no region choice is presented to the buyer as a selectable option, and nothing changes between the three published tiers on deployment.
AMENDED 6 September 2026 under R94 on newly located agreements; the pull-1 D recorded no deployment information. Regions are published and the AI processing location is disclosed candidly, short of a tenancy statement. The Platform Terms of Service effective 1 October 2025 are built around a Data Location the subscriber selects, with separate privacy policies for Australia, the United States, Europe and Canada; section 1.4 of the Artificial Intelligence Addendum states that AI processing occurs in data locations based on third-party AI tool availability and may differ from the subscriber's selected Data Location, with the applicable processing location displayed in the Subscriber Account. Whether customers share infrastructure is not stated. Terms of Service excerpts and AI Addendum read 6 September 2026.
Security Certifications and Trust Center
Independent attestation a buyer can pull without a sales call: SOC 2, ISO 27001, penetration test summaries, a trust center with current reports and named scope rather than a badge image.
Two certifications are stated with an audit cadence, and the evidence behind them is not reachable. ISO 27001 controls and the wider Information Security Management System are stated to be internally and externally audited annually, and SOC 2 internal and external audits are said to be undertaken annually with reports produced. That is more than a badge wall and it is why this sits here. Four things hold it below the top band. No auditor or certifying body is named for either. No SOC 2 type is stated, so a reader cannot tell whether the report is a point-in-time Type 1 or a period Type 2. No report date, period or scope statement appears, and no route to obtain either report is published, with no trust portal and no request form. And the ISO claim names the superseded edition, citing ISO 27001:2013 on a page last modified 21 April 2026, after the October 2025 deadline for transition to ISO 27001:2022. Recorded because a reader who checks the standard will find it withdrawn. The AWS certifications listed alongside, including FedRAMP and SOC 1, are correctly attributed by the vendor to AWS rather than claimed, and are not credited here.
Certification is real, named and consistently stated across the property, short of scope and evidence detail. ISO 27001, SOC 2 Type 2 and HIPAA are all named, described by the vendor as third party accreditations, and repeated identically on the security page, the FAQ, the pricing page and each product page, which is more internal consistency than several records here manage. The vendor states it undertakes regular comprehensive independent audits of its applications, systems and networks, and publishes a dedicated information security page explaining what each standard covers rather than only displaying badges. Security documentation is available on request and the SOC 2 report is obtainable by contacting the team, which is a request flow rather than a sales gate. Short of an A because no coverage period, audit scope, report date or auditing firm was located for any of the three, and no trust portal exists, the route being a request to the vendor.
Model Supply Chain Disclosure
Which models sit underneath, whose they are, where they run, and whether the vendor commits to telling customers when that changes. A legal buyer inherits every dependency it cannot see.
The architecture is described in useful detail and the models and their makers are not identified. What is published is real: all AI tools run on Amazon Bedrock, each model runs in a dedicated AWS account, Bedrock does not store inputs or outputs, does not share them with third parties and does not use them for training, and inference runs in Australia or the United States according to the customer's location. A buyer can therefore establish where the model runs, on what service and under what data conditions. What is not published is whose model it is. Bedrock hosts models from several makers and none is named, no model or version is identified, and nothing states which model handles a summary as against a playbook comparison. No commitment to notify customers if the model, version or hosting arrangement changes was located. Amazon appears here as the inference service operator rather than as infrastructure, and its separate role hosting the application and database is graded on the deployment row rather than counted twice.
AMENDED 6 September 2026 under R94 on a newly located agreement; the pull-1 D recorded no disclosure. The vendor now acknowledges third-party models in the agreement without identifying them there. The Artificial Intelligence Addendum defines Third-Party AI Tools as the third-party language models and AI technology Xakia uses to deliver the AI features, states that Xakia has no control over their operation or continued availability, discloses that their processing location may differ from the subscriber's Data Location and is shown in the account, and provides at section 6 that any flow-down terms the providers require are displayed at trust.xakiatech.com; section 8.3 gives thirty days' notice of addendum amendments. No provider or model is named in the addendum and the trust centre was not opened, which is the rebuttal route. AI Addendum read in full 6 September 2026.
Commercial Transparency
Whether a buyer can learn what this costs without entering a sales process: published rates, the unit being charged, what sits behind an enterprise tier, and what implementation adds.
A buyer can price this completely without speaking to anyone, which is rare in this lane. Three tiers are published with rates: Launch at 105 per user per month, Scale at 130 and Optimize at 160, with a currency selector offering Australian dollars, US dollars, euros and pounds. The unit is stated plainly as per user per month. What implementation adds is published rather than withheld: no implementation fee on Launch, optional customised configuration and implementation at 5,000 on Scale, and the same 5,000 as a standard inclusion on Optimize, with reconfiguration priced as a professional services fee on the lower tiers and one annual reconfiguration included at the top. A full feature comparison table runs to roughly forty rows across matter management, documents, spend, collaboration, AI, workflow, reporting, integrations, security and support, marking each as included, excluded, quantified or coming soon. AI itself is tiered transparently, with usage caps on Launch and unlimited access above it. A fourteen-day free trial is offered with self-serve signup.
FIRST A ON THIS AXIS IN THE INDEX, and by a wide margin. A published pricing page sets out named tiers being Advance, Professional, Enterprise and All-In, states the unit of charge as per user per month, offers month to month plans with no lock in alongside discounted annual subscriptions, and provides a free 14 day trial requiring no credit card. Onboarding is disclosed as a separate one off fee with light and full options covering data migration and training, so implementation cost is surfaced rather than discovered later. AI is priced explicitly: an add on for Advance and Professional, included with Enterprise and All-In, so a buyer knows before contact whether the capability they want carries an extra charge. Two further disclosures go beyond what this axis requires and are recorded because they are rare: the vendor states outside counsel are never charged to use the Connect portal to submit invoices, and that it takes no percentage of the value of invoices processed, naming and rejecting the clip of the ticket model some competitors use. Not captured in this pass: the specific rates at each tier, which the pricing page presents but which were not retrieved here. The grade rests on the published structure, unit, trial terms and fee disclosures, which together let a buyer understand the commercial model without contacting anyone.
Firm and Practice Coverage
Who the product is actually built for. AmLaw, midlaw, small firm and solo, in house departments, government and courts, and which practice areas are supported rather than merely claimed.
The buyer is described with substance and the practice boundary is left open. The segment is stated repeatedly and consistently: in-house corporate legal departments, with the product characterised in the vendor's own strapline as a simple end-to-end matter management tool for smaller corporate legal teams, and the pricing tiers structured per user so that a team can size itself. Named customers support the claim at the mid-market and large end, with MYOB and Equity Trustees both Australian listed or substantial institutions. Coverage extends beyond legal in a way the record should carry: the same platform is sold to HR teams under a separate product line, and the vendor describes a corporate front door spanning legal, HR and finance. Practice area does not bite on a matter management product and is named rather than penalised, since the system is matter-type agnostic. What is missing is the limit: no jurisdictional coverage is stated despite Australian and United States hosting, no minimum or maximum team size is given, and nothing identifies work the product is not suited to.
Segment coverage is described with substance and includes an explicit self limit, which is rare on this index. The target is stated numerically rather than vaguely: in house legal teams of roughly 2 to 200, which tells a prospect outside that band to look elsewhere. Language coverage is enumerated and used as a differentiator, the platform being configurable in English, Japanese, Spanish and French with the vendor claiming it is the only multi lingual platform in its market, and the spend module supports all currencies for global teams. Functional scope is stated at module level across matters, intake, contracts, spend, documents, entities and reporting. Not located as of 29 Aug 2026: any industry segmentation, jurisdictional coverage stated as such, and any statement of which practice areas or work types the platform is not built for beyond the team size band.
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?
Public material states that customer content is not used for training, and there is no agreement in which a matching term could sit. Two FAQ entries carry it: information is never used to train external AI models, and the platform uses only the data within the specific document or matter being worked on combined with Amazon Bedrock's services, so data is not used to train external models. The security section adds that Bedrock does not use inputs or outputs for training.
The qualifier belongs on the record and is not incidental: every one of these statements is scoped to external models, and nothing published states whether Dazychain or Yarris uses customer content to train, tune or evaluate anything of their own. The agreement search this value requires was performed and returned nothing to search: the site inventory taken from the navigation and footer on 4 September 2026 contains a privacy policy and a sitemap and no terms of service, subscription agreement or end user terms, so no contractual term exists to check either way. The privacy policy was not opened in this pass.
AMENDED 6 September 2026 under R94 on a newly located agreement; the pull-1 value was silent because no agreement had been read. The published agreement permits training on de-identified data derived from customer content, alongside a bar on training with the content itself. The Artificial Intelligence Addendum effective 1 February 2026, which forms part of the Platform Terms of Service, provides at section 2.4 that Xakia will not train any AI model using Inputs or Outputs and that its third-party AI tools will not either, and that use of Inputs and Outputs to improve service features does not constitute training; section 2.5 then permits Xakia to collect Derived Data, defined as anonymous data and statistics derived from Inputs, and use it for benchmarking, reports and training the AI features or AI models, disclosing it externally only de-identified and aggregated.
That is an express contractual permission to train on anonymized derivatives of customer content, which the index records at this value under the R43 Clio precedent; the section 2.4 protections and the third-party bar are the qualifiers a buyer should weigh. Surfaces checked 6 September 2026.
Prompt and Output Retention
How long does the product keep what a lawyer typed, and can that be set to zero?
The model leg is stated as zero retention and the platform leg is set by the customer. On the AI path the security section states that Amazon Bedrock does not store inputs or outputs, does not share them with third parties and does not use them for training, with each model running in a dedicated AWS account, so nothing is described as persisting from a prompt or a generated summary. On the platform path the position is set by the customer rather than by a fixed period: on account termination customer data is retained for the period the customer requests, usually one month, after which it is deleted, with alternative arrangements available on request.
What is not published is any retention position during the subscription itself, and nothing states how long a generated matter or document summary is held as part of the matter record once it has been written into it, which is the point at which model output becomes ordinary platform data.
AMENDED 6 September 2026 under R94 on a newly located agreement; the pull-1 value was not-addressed. A specific period is published for AI prompts and outputs, and it is zero: section 2.8 of the Artificial Intelligence Addendum states that because of the ephemeral processing of the AI features Xakia does not retain Inputs or Outputs after the Output is delivered, that Inputs are processed transiently and not stored, and that the customer is responsible for capturing anything it needs to keep.
The period is fixed by the vendor rather than configured by the customer, and it governs the AI features rather than matter data held in the platform, whose retention sits in the Terms of Service. Surfaces checked 6 September 2026.
Ethical Walls and Matter Segregation
Does retrieval respect the firm’s ethical walls, or can the model read across them?
Separation between customers is documented and separation within a customer is not. The security page states that data segregation is fully supported and tenant aware, that no user company can access another's information unless permission has been provided as in a collaboration, and that role-based access applies with every access and modification logged. On the AI side the boundary is described more tightly than most: the model reaches nothing until a user engages a function, and it works only within the matter or document it was engaged on rather than across the account.
What is absent is the matter-level wall inside a single legal department. Nothing describes conflicts management, restricted matters or a walled-user view, and the vendor's own FAQ asks whether AI can be restricted by role or matter type and answers only that it can be switched off for the whole account, which leaves the finer half of its own question unanswered.
One of the clearest segregation disclosures on this index for an in house product, and materially better than the category norm. The vendor publishes role based permissions controlling access at individual user, team or matter level, states directly that sensitive matters can be restricted to specific individuals, and states that external parties such as outside counsel see only what the customer chooses to share with them through the Connect portal.
That is matter level restriction described as a product capability with a named use case, plus a bounded external party view, which together address both halves of the question this signal asks for a corporate legal buyer. Single sign on through the customer's own identity provider governs authentication. Recorded at own model documented rather than the positive value because the product operates its own permission structure rather than inheriting a document management system's access model at query time, and because nothing located states that the AI features respect 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?
Nothing located addresses compelled disclosure or customer notice. The evidence home for this signal is the confidentiality section of a customer agreement, and no agreement is published: the site inventory taken on 4 September 2026 carries a privacy policy and a sitemap and nothing else. The nearest published statement runs to voluntary sharing rather than compulsion, with the security page stating that data is not shared with third parties and that users may choose to share information with external lawyers if they wish.
No transparency report, government request statement, law enforcement section or notice commitment appears on the security page, the AI page, the pricing page or the features index. The question has weight here because hosting spans Australia and the United States, so a customer's matter data sits under two disclosure regimes and nothing published says what happens if either is invoked. The privacy policy was not opened in this pass and may address it.
Searched the security page, the FAQ, the pricing page and the site footer on 29 Aug 2026, and no published customer agreement, terms of service or data processing agreement was reached. No clause committing to notify a customer of a government or law enforcement request for their data was located, and no transparency report was located. This records a search of the public pages rather than a reading of contract documents, none of which were located as published.
Primary Law Corpus Provenance
Where does the law in this product come from, and does the vendor have the right to use it?
The inputs are named and they are entirely the customer's own. Published material states that the AI uses only the data within the specific document or matter being worked on, and the features described operate on uploaded contracts, matter records, emails and the customer's own contract playbook. No external corpus is involved: the product does not retrieve primary law, published precedent, a market clause bank or any licensed dataset, and none is named anywhere.
The one component that could raise the question is the contract playbook comparison, and the playbook is expressly the customer's own corporate document rather than vendor-supplied content. There is accordingly no licensing question of the kind this signal was written for. Searched the AI page, the features index, the security page and the pricing page on 4 September 2026.
No primary law corpus is identified because the product does not hold one. The AI features operate on the customer's own contracts, documents and invoices held in the platform, so the corpus is the customer's own material and its provenance is theirs. Searched the platform pages, the FAQ and the in house hub on 29 Aug 2026 and located no vendor supplied legal corpus, no license basis and no update cadence, and none would be expected for a matter management product of this shape. Same architectural position as the other legal operations records on this index.
Good Law Verification
Does the product tell you when the authority it just cited has been overruled?
Nothing on any located surface addresses checking authority for subsequent history, and the product does not retrieve or present primary law. Dazychain manages matters, documents and spend and its AI summarizes, searches, answers questions about the customer's own documents and compares a contract to the customer's playbook; no case, statute or regulation is surfaced to a user at any point in the published workflow. The question does not bite on this product class and the value records the honest absence rather than a shortcoming. Searched the AI page, the features index, the security page and the pricing page on 4 September 2026.
Searched the platform pages, the FAQ and the in house hub on 29 Aug 2026. No material was located addressing whether authority carries a treatment signal or whether subsequent history is checked, and no commercial citator license was located. Noted for context: this is a matter management and legal operations platform whose corpus is the customer's own matters, contracts and invoices rather than published case law, so a citator is outside its design entirely.
Refusal and Uncertainty Behavior
What does the product do when the answer is not in the corpus?
A published limitations statement exists and it addresses scope and input quality rather than uncertainty. The FAQ entry headed on limitations and best practices states that the AI works only within the matter or document it is engaged on and does not access data outside that context, and that clear, well-structured and complete documents lead to better results. A second entry states that extractions are designed to be highly accurate but may require review, and that users can validate and edit them in the platform before anything is finalized or shared.
Both are recorded here as what exists, and both are more candid than most vendors offer. Neither describes what the system does when it cannot ground an output. No confidence or certainty score is surfaced against an extraction, no abstention or no-answer state is described, and nothing addresses the ordinary failure conditions for this product class, such as a scanned or hand-amended contract, a clause spanning pages, or a playbook comparison where no corresponding clause exists.
Searched the platform pages, the FAQ, the security page and the product updates section on 29 Aug 2026. No published material describes what the AI features do when they cannot ground an answer, and no explicit no answer path or confidence signal exposed to the user was located. The vendor's statement that it expands AI capabilities without compromising accuracy is an assertion about quality rather than a description of behavior under uncertainty, and the two were not conflated.
For key terms extraction in particular, what the product does when a term is absent or ambiguous is a live question and is unaddressed.
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 was searched on 4 September 2026 on the product names Dazychain and IntuityAI and on the parent company name Yarris. No court order, opinion or disciplinary record naming any of them was located. This records the state of the public record on that date and is not a finding about the product. The signal also sits at an angle to this product class, since the AI summarizes and answers questions about the customer's own matter documents rather than generating legal citations, so a fabricated citation is not the failure mode it would ordinarily produce.
No court order, opinion or disciplinary record naming this product has been located as of 29 Aug 2026. Instrument searched: the AI Hallucination Cases database maintained by Damien Charlotin, which tracks court decisions worldwide involving AI generated hallucinated content and records the AI tool implicated where it is known. Also checked published 2026 sanctions summaries and secondary sanctions trackers. The entries located name filers, and in some rows other products, rather than this one.
This is a statement about the public record on the date shown and not a clearance. Note the product generates contract redlines, summaries and extracted terms from the customer's own documents rather than citations to authority, so the failure mode this database catalogs does not arise directly, and note the vendor is Australian headquartered while the database is weighted toward United States filings.
Bar Guidance Alignment
Has the vendor engaged in public with the ethics opinions its buyers are bound by?
No bar authority, regulator, conduct rule or ethics opinion is named on any located surface. The AI page, the security page, the features index and the pricing page were read on 4 September 2026 and none engages professional regulation, and no jurisdiction-specific guidance is mapped. The compliance material that does exist is data protection rather than professional conduct, running to the Australian Privacy Act 1988 and related state laws, GDPR, ISO 27001 and SOC 2.
The absence is worth noting against one published feature in particular: business stakeholders who are not lawyers can complete an intake form that automatically generates a finished document such as a non-disclosure agreement from a pre-approved template, and nothing published engages the professional responsibility question that raises for the legal department that approved the template.
Searched the site, the FAQ, the in house hub articles and the product updates section on 29 Aug 2026. No engagement with any named ethics opinion or professional guidance was located, including ABA Formal Opinion 512, United States state bar guidance, and Law Council of Australia or state law society guidance given the vendor's home market. The vendor publishes practical guidance for in house teams on evaluating matter management software and on legal operations practice, which addresses procurement and process rather than the professional conduct obligations its users are bound by.
Billing and Fee Posture
Does the vendor address what happens to the bill when the work takes an hour instead of six?
Time-saving claims are published and the billing question is not reached. The AI material states that the platform helps lawyers save time, removes friction from drafting, reporting and summarization, and enables teams to get work done in a fraction of the time, without attaching a figure to any of it. Nothing addresses what happens to a bill when that work compresses. The product does carry substantial billing machinery, with spend and budget management, invoice review workflows, spend analytics, schedules of rates support and budget tracking, but all of it concerns what the department pays its external law firms rather than any record of AI-assisted work, and the established treatment is that generic cost tooling of that kind does not answer this signal.
No per-matter record of AI involvement is described as available, and nothing states whether a matter summary or a playbook comparison produced by the model is identified as such in the matter record or in any report to the business.
The product holds the raw material and the vendor's published position runs to its own fees rather than to the client's. Spend and budget management tracks legal spend in real time with invoice review among the AI features, and the Connect portal carries invoice submission from outside counsel, so a legal department using the platform holds a structured per matter record of what firms billed. The vendor's distinctive disclosure is about its own charging: outside counsel are never charged to use the portal, there is no limit on invoices received, and no percentage is taken on the value of invoices processed, which it names and rejects as a clip of the ticket.
That is a vendor being explicit that it does not profit from the size of its customer's legal spend, and no other record on this index makes that statement. Searched the platform pages, the pricing page and the in house hub on 29 Aug 2026 and located no per matter record of AI assisted work intended for fee purposes and no guidance on billing, fee or client disclosure treatment.
Outside Counsel Guideline Readiness
Can a firm get this vendor through a client’s AI clause without a bespoke negotiation?
Infrastructure is named, the model provider is not, and there is no forwardable pack. What a buyer can establish from the security and AI pages is real: application and file hosting with Amazon in Australia and the United States, the database with Object Rocket in the same regions, AI inference on Amazon Bedrock with each model in a dedicated AWS account and no storage of inputs or outputs. But naming a cloud host and a database provider says where the platform runs rather than whose model reads a contract, and infrastructure alone does not satisfy this signal.
Bedrock hosts models from several makers and none is identified, so the question a counterparty would actually ask cannot be answered from published material. There is no subprocessor register, no data processing addendum, no consent or notification pack, and no artifact drafted to be forwarded. The direction of the signal also inverts on an in-house product, since the buyer is the client rather than the firm, and that is recorded rather than treated as mitigation.
A stated route to diligence material exists and the specific artifacts this signal names do not. The vendor publishes that security documentation can be requested at any time and that the SOC 2 report is obtainable by contacting the team, alongside a dedicated information security page naming ISO 27001, SOC 2 Type 2 and HIPAA and explaining what each covers. Matter level access restriction and the bounded outside counsel view through Connect are genuine answers to part of what a client AI clause asks, since they bear on who can see client material.
Searched the security page, the FAQ and the pricing page on 29 Aug 2026 and located no subprocessor list, no statement naming which model providers see customer content, no published data processing agreement, and no client facing consent or notification material. Recorded at on request on the strength of the stated documentation route.
Court Disclosure Support
If a judge’s standing order requires an AI disclosure, can the product produce one?
A comprehensive activity record is published and nothing states that it distinguishes model work from human work. The security page describes every access to the platform and every modification to legal documents, matters and deliverables as logged and preserved, framed as a safeguard against unauthorised activity, and audit trails and activity tracking appear as an included feature at all three pricing tiers alongside exportable reports and dashboards.
That is a real and exportable record of what happened to a matter, which is why this sits above the floor. The AI-specific half is missing entirely. Nothing says the log records which summaries, extractions or playbook comparisons were produced by the model rather than entered by a person, no model or version is attributed to any output, and no guidance or template for disclosing AI use to a court, regulator or client is published.
Searched the platform pages, the FAQ, the security page and the in house hub on 29 Aug 2026. No per document record covering model used, sources retrieved and human verification was located, and no model is identified in published material so the model used could not be stated. The platform records matter activity, intake routing and spend against each matter, so a workflow trail plausibly exists, but nothing published describes an export or a defensibility record for AI generated output specifically.
Noted for context: this is an in house matter management platform whose output is internal records, contracts and reporting rather than court filings, so a judicial AI disclosure order 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.
- AI Governance and Bias Disclosure
- AI Liability and Recourse
- Third Party Request and Subpoena Notice
- Primary Law Corpus Provenance
- Good Law Verification
- Refusal and Uncertainty Behavior
- Bar Guidance Alignment
Which one fits
Choose Dazychain if
- You want AI that does nothing until someone asks. Dazychain's AI runs only when a user engages a function, works only inside the matter or document it was opened on, and can be turned off entirely for an account.
- You want to know where the model runs and what it keeps. Dazychain runs its AI on Amazon Bedrock in Australia or the United States by customer location, with each model in a dedicated AWS account, and states that inputs and outputs are not stored or used to train external models.
- You want to show the business what legal is carrying. Dazychain's dashboards report matter volume, workload and lifecycle by team, function or region, and MYOB's deputy general counsel describes using that data to argue successfully for more headcount.
Choose Xakia if
- You want the AI's limits written into the contract. Xakia's AI addendum states that output is general information rather than legal advice, that the features must be used by or under the direct supervision of qualified lawyers, and bars automated decisions with legal effect without proper human review.
- You want AI prompts and outputs gone once delivered. Xakia's AI addendum states that inputs are processed transiently and that it does not retain inputs or outputs after output is delivered, and it bars its third party AI providers from training on them.
- You work in several languages or currencies. Xakia can be configured in English, Japanese, Spanish and French, its spend module supports all currencies, and outside counsel submit invoices through its Connect portal at no charge, with no percentage taken on invoice value.
In summary
Dazychain
Dazychain, made by Yarris Technologies of Melbourne, is matter management software for in house legal departments: requests arrive through a front door form or email and become matters with workflows, tasks and approvals, alongside document and contract management, portals for business stakeholders and law firms, spend tracking and dashboards. Its AI, DazychainAI, searches, summarizes matters and documents, answers questions and compares contracts to a playbook, only when a user engages it. The AI Legal Index grades it in the top two bands on ten of fifteen capability axes, with an A on pricing, published at three per user tiers. It names MYOB and Equity Trustees among customers. As of 4 September 2026 the index located no customer agreement, named model or breach notice commitment.
Xakia
Xakia, founded in 2015 in Melbourne, is a matter management and legal operations platform for in house legal teams of roughly 2 to 200, covering matters, intake and triage, contracts, spend and budgets, documents, entities and reporting, with the Connect portal for outside counsel. Its AI features review and redline contracts, summarize them, extract key terms, search and review invoices, as an add on on lower tiers. The AI Legal Index grades it in the top two bands on six of fifteen capability axes, with A grades on professional responsibility and pricing, from $95 per user a month. Its AI addendum draws the advice line and states zero retention of AI inputs. As of 29 August 2026 the index located no named model, accuracy measure or liability position.
Questions buyers ask
Dazychain vs Xakia: which is better for a small in house legal team?
On published evidence Dazychain sits in the top two bands on ten of fifteen AI Legal Index capability axes and Xakia on six of fifteen, identical on nine. Dazychain bounds its AI more tightly and publishes where it runs. Xakia puts its AI limits in a contract addendum and supports four interface languages. Both publish per user prices, so a team can compare costs before a demo.
Does Xakia train AI on customer data?
Not on the data itself. Xakia's AI addendum, effective 1 February 2026, states that neither Xakia nor its third party AI tools will train any AI model using inputs or outputs. It also permits Xakia to collect anonymous data and statistics derived from inputs and use them for benchmarking, reports and training its AI features, disclosing them externally only in aggregated form. Graded by AI Legal Index against 15 capability axes and 12 legal signals, including privilege handling and citation accuracy, from each vendor's own published materials, verified September 26, 2026. No vendor pays for placement.
Can Dazychain's AI be turned off?
Yes. Dazychain's FAQ states that AI can be completely turned off for an account, that it accesses data only when a user actively engages a function such as clicking a button, and that it works only inside the matter or document it was engaged on. The FAQ does not say whether AI can be restricted by role or matter type rather than for the whole account. Graded by AI Legal Index against 15 capability axes and 12 legal signals, including privilege handling and citation accuracy, from each vendor's own published materials, verified September 26, 2026. No vendor pays for placement.
How much do Dazychain and Xakia cost?
Dazychain publishes Launch at 105, Scale at 130 and Optimize at 160 per user a month, with a currency selector, implementation of 5,000 on the upper tiers and a 14 day trial. Xakia's Advance, Professional and Enterprise tiers run from $95, $120 and $150 per user a month, month to month or annual, with AI an add on below Enterprise and a 14 day trial. Graded by AI Legal Index against 15 capability axes and 12 legal signals, including privilege handling and citation accuracy, from each vendor's own published materials, verified September 26, 2026. No vendor pays for placement.
What do Dazychain and Xakia both leave unpublished?
Who bears the loss and how their AI is governed. Neither publishes a liability, warranty or indemnity position for its AI output, and neither publishes an AI governance framework, testing before release or any finding on uneven output. Neither measures how accurate its summaries or extractions are, and neither names bar guidance on AI. Graded by AI Legal Index against 15 capability axes and 12 legal signals, including privilege handling and citation accuracy, from each vendor's own published materials, verified September 26, 2026. No vendor pays for placement.
Three readings to weigh. Xakia's addendum permits training AI on anonymized data and statistics derived from customer inputs, and its AI may be processed in a location other than the one the customer selects, shown in the account; both are published terms. Dazychain publishes no customer agreement, so its AI commitments are policy rather than contract, and its ISO 27001 claim cites the superseded 2013 edition. Dazychain's price currency is set by a selector on its page. Dazychain was verified on 4 September 2026 and Xakia on 29 August 2026. Neither vendor reviewed this page.
Neither vendor paid for inclusion, placement or a grade, and neither reviewed this page before it published. Everything above comes from public material on the dates shown. How the index grades.