Hona vs Trailmate: how they compare in 2026
Hona's Lia is organized around the treatment phase of a personal injury case. It coordinates appointments, gathers symptoms and provider information after visits, and alerts the firm when treatment gaps or other issues emerge. Trailmate applies an assignment model across a wider set of client tasks and practice areas, from intake to discovery facts and closing communication. The AI Legal Index identifies Lia's repeated treatment follow up as a focused workflow connecting a client, the firm and medical providers. Trailmate lets staff define the work, its deadline and escalation points, then track completion. The AI Legal Index identifies those reusable assignments as the basis for adapting outreach to different matters. Both operate proactively and return exceptions to people. The practical difference is between a treatment centered agent and a general client communication agent configured for the firm's task.
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.
Hona has repositioned around the model, but the platform underneath still stands. The company now markets itself as AI for personal injury firms and Lia as an AI case manager, and the FAQ confirms Lia is priced and contracted separately from what the vendor calls Hona's core platform. That core (client portal, case progress updates, two way messaging, drip campaigns, branding and education) is conventional communication software that predates Lia and is still sold. Remove the models and a working client update product remains; the models power a core capability on top of it.
The agent converts an ordinary language assignment into a structured conversation and pursues the information needed to complete it. It can interview a prospective client, request missing records, clarify an incomplete answer and continue outreach as a matter progresses. Text and voice interaction, Spanish support and automatic processing of uploaded files make the model driven conversation the core service. Deadlines, reminder settings and reusable workflows constrain execution rather than replacing that intelligence. Without the models, the product would lose its ability to interpret delegated objectives and adapt client interactions, leaving only the surrounding records and notification controls.
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 through operating metrics and never measured as accuracy. The site publishes a 94% check in rate, a response time under four minutes, and a claim that 100% of treatment updates are captured automatically. None of those says how often Lia records a symptom, provider or appointment correctly, and no evaluation, test set or error rate is published on any page read. That matters because the extraction is the product: Lia reads free text client replies and writes structured treatment records that later support a settlement demand, so a misread provider name or date carries into the case file. Tests about primary authority and citators do not apply, since Lia is expressly barred from making legal assertions.
Trailmate collects information directly from clients and preserves session transcripts, attachments and generated updates for firm review. Voice conversations can use confirmation or repetition for critical details, while uploaded files may undergo optical character recognition and classification. The firm receives a summary of attempted work, completed tasks and information still requiring attention. No measured transcription evaluation, field extraction test or answer level link to supporting material is supplied. The terms warn that generated content can be incomplete or inaccurate and require independent review. The workflow gathers case facts rather than verifying legal citations against primary authorities.
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 control structure is complete and published. What runs alone is stated: Lia handles routine treatment phase communication herself, opening contact by text and running check in messages without approval of each one. What constrains her is written by the customer and specific: escalation rules in plain language, with the vendor's own examples being alerts when a client mentions stopping treatment or when a provider appears who is not on the preferred list. Thresholds are published: physical therapy missed twice in a row, no response for 48 hours, and high priority issues repeating to the case manager every hour until resolved. Review is built in: every conversation is visible to the team, staff can step in at any time, each case manager sees their own cases, and drafts to providers need human sign off before anything goes out. The route back to a person covers the highest risk case: a legal question goes immediately to the assigned case manager, and a call placed from the dashboard is transcribed back into Lia. Enrollment is optional case by case, and the vendor advises firms to exclude clients unlikely to respond well to text.
Firms define the assignment, deadlines, escalation points and intensity of reminders. An agent can be configured to pause when a person begins sharing privileged details before an engagement and redirect that person to an attorney. Progress updates distinguish completed work from matters requiring staff attention, and session transcripts remain available for monitoring and export. Trial and demonstration access is restricted to evaluation outside production. These controls place boundaries around delegated work, but the product does not specify an error recovery procedure or a universal stopping threshold. Lawyers remain responsible for validating collected facts and generated summaries before relying on them.
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 users at named firms, and figures that belong to the platform rather than any of them. People on record with titles and employers include Sofia Escobedo, Assistant at The James Law Firm; Crisely Gonzalez, Case Manager at Omega Law Group; Alexandria Zant at Zant Law Firm; Domenic Martini at Singleton Schreiber; and Bridgett Sanne at Inserra Kelley Cooper Sewell. Firm logos include Isaacs & Isaacs and Nachawati Law. Separately, the vendor claims more than 500 personal injury firms, a 94% average client intake completion rate and an average response time under four minutes. None of the quoted users carries a figure, and none of the figures carries a customer, date or method. The market statistics on the product page, a 30% reduction in payouts where treatment gaps occur and more than 20% of cases showing a two week gap in the first six weeks, describe the industry, not Hona.
Founder Manny Starr describes developing Trailmate within an operating law firm, and the company's account of Frontier Law Center explains the client communication problems behind that development. The work involves intake, repeated document requests and coordination that continued despite improvements to internal drafting and research. The careers material also describes a live product used by firms and clients. These accounts identify an operational setting rather than a purely proposed concept. The current product materials do not provide a measured customer outcome with dates, a baseline and a calculation method. Illustrative completion figures inside the agreement are examples of permitted marketing disclosures, not performance results.
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.
The agreement is published and substantive on confidentiality, and silent on training. Terms clause 8 sets a mutual confidentiality regime with a three year tail and perpetual protection for trade secrets. Clause 13.4(b) requires Provider to stop all use of Customer Data on termination and return or destroy it, certified in a signed instrument. The compliance section states PHI is handled under a BAA with end to end encryption, and clause 9.2 leaves the customer owning Customer Data. But nothing addresses training on client content, and clause 9.1 runs the other way: the customer unconditionally and irrevocably assigns to Provider all right, title and interest in Resultant Data, a term defined only in a separate Definitions Exhibit that was not read. No matter or tenant segregation is documented, and privilege and work product are not addressed for a system carrying communications between a firm and its clients.
The agreement treats customer material, including privileged communications and work product, as confidential. It limits access to authorized people with a need to know and confidentiality obligations, and prohibits customer data from training foundational or other general purpose models. Operational improvement and anonymous usage analysis remain permitted. Firms control who joins a session and can configure data collection boundaries. The customer retains responsibility for preserving privilege and obtaining consent. External model providers are not identified, and their handling of inference requests is not separately explained. The public commitments therefore do not resolve every downstream access and retention question.
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.
Two published commitments stand out. First, the advice line is a product guardrail, not a disclaimer: Lia is set never to give legal advice or legal opinions, and any question about legal strategy goes straight to the assigned case manager. Second, the vendor commits to telling the client about the AI. Lia introduces herself as an AI from the firm in her first message, described as a requirement rather than an option, with the recommended practice that a person sends an introduction first. For a client who cannot tell whether an answer is legal advice, that is the disclosure that matters. No jurisdiction limits are named, nothing addresses the firm's own supervision and competence duties beyond the escalation route, and ABA Formal Opinion 512 is not mentioned on any page read.
Trailmate serves legal practices and their authorized staff, with clients and witnesses participating only through the firm's invitation. The agreement states that it supplies software rather than legal advice or representation, and that no attorney client relationship arises with Trailmate. Lawyers retain independent judgment, output verification, confidentiality obligations and responsibility for competent representation. Product guidance supplies an introductory AI disclosure and describes routing sensitive conversations back to an attorney. Clients receive firm controlled instructions about the purpose of the interaction and its administrative scope. A jurisdiction specific map of recording consent, professional rules and permitted client interactions is not provided.
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 is published for a system that talks to injured claimants unsupervised. The home page, the AI case agents product page with its compliance and FAQ sections, the privacy policy and the full Terms and Conditions were read. There is no responsible AI page, named owner accountable for the AI, testing regime before release, or ISO 42001 or equivalent. Nothing addresses whether Lia performs evenly across client groups. That is a live question for a product that runs in English and Spanish and whose vendor advises firms to exclude elderly clients and catastrophic injury cases from enrollment, an exclusion that implies limits it never describes. The legal advice guardrail and escalation rules are oversight controls, and TCPA and HIPAA compliance are regulatory posture, not model governance.
The product restricts the agent to assignments and data boundaries defined by the firm, with escalation when a conversation requires professional judgment. Its privacy guidance promotes transparent AI disclosure, limited collection and retention policies appropriate to the matter. Those principles have practical customer configuration mechanisms, including a pause for sensitive information before engagement. They are not accompanied by a vendor model governance framework with an accountable owner, release approval criteria or test results. No assessment describes uneven performance across Spanish and English, different client populations or practice types. Named company leaders are not assigned an explicit AI risk oversight role.
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.
Deletion is well handled and little else is. Terms clause 13.4 requires Provider, on termination, to stop using Customer Data and return or destroy it, certified in a signed written instrument, with carve outs for legal retention and for backups, archives and disaster recovery systems until deleted in the ordinary course. The compliance section states end to end encryption and PHI handled under a BAA, and clause 5 commits to routine backups, with restoration as the sole remedy for data loss. Missing are a retention period during the term, any incident or breach notification practice, and a subprocessor list. Clause 1.6 lets Provider engage subcontractors at its discretion with no list and no duty to give notice. Clause 6 defines the security commitment by pointing to the privacy policy, dated 12 September 2022, which predates Lia. The policy covers the product without addressing what happens to client messages and generated records after processing.
Encryption, authentication, access controls and logging protect information handled by the service. Firms configure roles and session invitations, and external participants can use a passcode or secure link. The agreement requires incident notice without unreasonable delay, investigation and reasonable cooperation. Firms may set retention policies; on termination, a 30 day export period precedes deletion from active production systems unless earlier deletion is requested. Backups, legal obligations and claims can justify retained copies under confidentiality. A product subprocessor list, inference log expiry and a defined backup deletion schedule are not available alongside these controls.
AI Liability and Recourse
What the vendor stands behind contractually when its output is wrong. Indemnities, caps, carve outs, insurance, and whether any of it is published or only reachable through a negotiated agreement.
A complete, readable allocation is openly published. Terms clause 12.2 caps total liability at fees paid in the twelve months before the act giving rise to it, and clause 12.1 excludes the usual heads of consequential loss. Clause 11.1 gives the customer a real indemnity: Provider defends and holds the customer harmless against third party US intellectual property claims arising from use of the Services, with itemized carve outs. Clause 11.4 obliges Provider to procure rights, modify or terminate if the Services are enjoined. Clause 10.2 warrants performance by skilled personnel in a professional and workmanlike manner to generally recognized industry standards, and clause 4 adds a 95% availability commitment with service credits. None of it reaches output: clause 10.4 warrants nothing about results being accurate, complete or error free, the indemnity covers infringement rather than a wrong treatment record, and no insurance position is published.
Trailmate warrants professional service and substantial conformity with its product instructions. Its remedy is reasonable correction effort or, if correction fails, termination with a proportional refund of unused prepaid fees. A separate vendor indemnity covers specified intellectual property claims against the service, with exclusions for customer data, modifications and combinations with other systems. The general liability cap is six months of fees, with carve outs including confidentiality, indemnity, fraud and willful misconduct. Generated output accuracy is expressly disclaimed, and the firm must validate it. No specific compensation undertaking covers a mistaken client communication or an incorrect case fact.
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.
Integrations are numerous and aimed at the systems this buyer actually runs. Fifteen platforms are named on the product page, including Filevine, Clio, MyCase, CASEpeer, SmartAdvocate, Assembly, Smokeball, Lawcus, Salesforce, Lead Docket, Cerenade, Lexamica and AttorneyShare, which is the plaintiff side stack rather than a generic connector list. Direction is stated broadly: Lia sits alongside the case management system and pulls data both ways. The mechanism is named: enrollment happens through a Chrome extension that gives Lia the case context, and can be triggered by case status changes in the underlying system. No page read says which fields move, in which direction, or what a firm must configure, and the FAQ refers specifics to a sales rep. The dedicated integrations page was not opened and may say more.
The firm can embed Trailmate on its website and initiate client contact through email or text. Staff can export session transcripts and collected files, preserving a route for moving the resulting information elsewhere. The agreement contemplates optional connections to case management, storage, telephony and signature services, with configuration responsibilities assigned to the subscriber. It does not identify a current named connector or explain its field mapping, synchronization direction and supported actions. Generic contractual permission for integrations does not establish an available connection to a particular practice platform. The current workflow therefore has explicit export support without a verifiable practice system integration.
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.
Cloud delivery is evident, and neither tenancy nor region is stated. No tenancy model is published and no region option is offered. The only location material is in the privacy policy of 12 September 2022. It says personal data may be stored and processed in any country where Hona or its service providers do business, including the United States, that officials there may be entitled to access it, and that EU Standard Contractual Clauses are used where required for transfers out of the EEA, UK or Switzerland. That is a transfer disclosure, not a residency offering, and it fixes no processing location. Nothing separates where data is processed from where it is stored.
The service runs as a hosted platform with an organizational environment for each subscribing firm. Authorized staff receive their own access credentials, and external participants enter selected sessions through configured links or authentication. Client interaction can be embedded on the firm's website rather than requiring a separate installed application. The platform's hosting provider, storage region and inference location are unspecified. No private cloud or customer premises option accompanies the hosted service description. The company's Los Angeles contact address and Delaware incorporation identify the business, not the location where conversation audio, attachments or generated updates are stored and processed.
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.
The compliance section of the product page states SOC 2 Type II audited annually, with the report available on request. It also states HIPAA, with PHI handled under a BAA and end to end encryption, and TCPA, with consent tracked for each outreach and opt outs honored, which is the right standard for a product built on automated texting to consumers. Full documentation is said to be available on request. Nothing lets a buyer test any of it: no auditor, scope, trust services criteria, coverage period or report date is given, there is no trust center or portal, and the request route is stated but not described. The standard is named in prose with an annual cadence, not just as a badge.
Contractual security measures include encryption, authentication, access controls, network safeguards, security assessments and employee training. Incident response obligations cover notice, investigation and mitigation after unauthorized access becomes known. Those commitments describe the service's responsibilities but do not name an independent security certification or examination. No SOC 2 report, ISO certificate, audit period or scoped assessment is available with the product and legal materials. The privacy guidance discusses qualities a secure legal AI system should have without identifying a completed certification for Trailmate. A public report request process or trust center is also unspecified.
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.
Nothing is published about the model supply chain a customer inherits. Hona markets an AI agent that texts and calls clients, transcribes calls, extracts structured medical information and works in two languages. Yet across the home page, the AI case agents page with its compliance and FAQ sections, the privacy policy and the full Terms and Conditions read, no model, model provider, inference location or change notice commitment appears. There is no subprocessor list, and the agreement runs the other way: clause 1.6 lets Provider engage subcontractors at its sole discretion, with no duty to name them or notify the customer. No reference to an underlying model layer is published at all.
Voice sessions combine automatic speech recognition with generated speech, while text sessions provide the same structured client interaction without audio. Uploaded files can undergo optical character recognition, classification and tagging before the firm reviews them. The conversation agent operates within the assigned task and produces transcripts and progress updates. This outlines several processing components, but the underlying model developers and versions are unnamed. No mapping identifies which provider receives recordings, attachments or conversation text, or where those steps run. The agreement permits service changes without a model specific advance notification commitment and does not supply a current AI provider register.
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.
There is no pricing page, yet the charging structure is published in detail in the contract. The site's Pricing link leads to a demo booking, and the FAQ answers pricing and contract term questions by referring buyers to a sales rep. The Terms and Conditions publish the whole model. Fees are Recurring and One-Time Fees against an Initial User Count and a Case Tier Limit, with Excess User Fees, Case Overage Fees and Text Overage Fees defined. The text allowance is a formula, the Case Tier Limit multiplied by fifty, with the vendor's example of 12,500 texts for a 250 case tier over twelve months. One real rate is published: 1.5 cents per text beyond the allowance. Fee increases need thirty days' notice, and renewal is annual and automatic. The unit and structure are stated, but no plan price is published at any level.
The service is sold through a negotiated sales order or an online plan accepted at checkout. The agreement provides for advance monthly or annual billing, automatic renewal and possible overages if a chosen plan limits users, matters or sessions. It does not select a definite charging unit or provide a public table of current plans, included volumes and prices. Product pages direct firms to a demonstration. Implementation charges, committed minimums and the cost of a particular level of client outreach are unspecified. General billing provisions explain contract administration without establishing the commercial shape of an available package.
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 focus is stated sharply. Hona describes itself as built for personal injury firms, repeats it in the site footer, and writes its product page entirely around the treatment phase of a PI case: providers, appointments, treatment gaps and symptom capture. Eight practice types have their own pages: personal injury, workers compensation, bankruptcy, mass tort, employment, criminal, immigration and family law. The vendor also states real limits. Enrollment is optional per client, and firms are advised to start with active treatment soft tissue cases and exclude clients unlikely to respond well to texts, naming elderly clients and catastrophic injury cases. Firm segmentation is missing: nothing addresses solo versus large firm, in house or government use, and the practice pages beyond personal injury are thinner than the core.
Trailmate addresses law firms with frequent client communication across intake, active matters and closing work. Dedicated workflows cover personal injury, employment, immigration, family law, criminal defense, landlord tenant matters, trusts and estates, lemon law, workers compensation and class action outreach. Immigration examples include application records and responses to information requests; workers compensation includes treatment updates and work restrictions. Spanish interaction broadens access for clients who prefer it. The agreement limits subscribers to legal businesses and authorized professionals. Specific jurisdiction exclusions, firm size limits and the full set of supported languages are not enumerated.
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?
No located term or policy addresses whether client content is used to train models. The full Terms and Conditions, the privacy policy of 12 September 2022, the home page and the AI case agents page including its compliance and FAQ sections were read on 1 September 2026, and none mentions training. One clause runs toward the vendor and a buyer should read it: clause 9.1 has the customer unconditionally and irrevocably assign to Provider all right, title and interest in Resultant Data, a term defined only in a separate Definitions Exhibit at hona.com/definitions that was not retrieved. Recorded as silent rather than permitted because no located term expressly reserves a right to train.
The customer agreement prohibits using Customer Data to train, develop or improve foundational models, large language models and other general purpose machine learning systems. Customer Data includes session transcripts, attachments, recordings and generated material. Anonymous Usage Data is separately defined and excludes Customer Data, and operational performance, interface and workflow improvements remain permitted.
The prohibition is therefore a contractual restriction with an explicit scope, not a blanket ban on every service improvement. External model providers and their separate processing terms are not named.
Prompt and Output Retention
How long does the product keep what a lawyer typed, and can that be set to zero?
Retention is acknowledged and no period is given. The privacy policy commits to keeping personal data as long as the services are provided and afterwards for legal and reporting obligations, without a period. The Terms are more specific at the end of the relationship than during it: clause 13.4 requires Provider on termination to cease all use of Customer Data and return or destroy it, certified in a signed written instrument, while permitting retention where required by law and in backups, archives and disaster recovery systems until deleted in the ordinary course.
Nothing states how long client messages, transcripts or generated treatment records persist while the contract is live, and no customer control is offered.
Firms can set retention policies for client information and purge records when a case closes or a lead does not convert. The agreement also permits earlier deletion by written request and provides a 30 day export period after termination. Deletion from active production systems is subject to exceptions for legal obligations, claims, system logs and backups, with retained data still confidential. No zero retention setting or specific backup expiry is supplied. The export period describes departure from the service rather than the duration of an active conversation record.
Ethical Walls and Matter Segregation
Does retrieval respect the firm’s ethical walls, or can the model read across them?
Segregation is asserted at the level of what a user sees, with no published detail on enforcement. The product page states that each case manager sees their own book through per-case-manager dashboards, while also stating that every conversation Lia has is visible to the team and that staff can step into any conversation at any time, which are pulling in different directions and are not reconciled anywhere. No document read on 1 September 2026 describes an access model, a permission inheritance from the underlying case management system, or how any boundary is enforced when Lia retrieves case context through the Chrome extension.
Each firm has an organizational environment, and administrators configure roles, permissions and authorized users. The firm controls which outside participants can enter each session and verifies their identity before exposing sensitive material. Passcodes, secure links and optional credentialed access support that arrangement. These are Trailmate's own access controls rather than inherited permissions from a named document management system.
No procedure explains synchronization with an external ethical wall after files or conversations have entered the platform, or how later permission changes affect previously generated material.
Third Party Request and Subpoena Notice
If someone subpoenas the vendor for a firm’s data, does the firm hear about it first?
The agreement commits to notice before disclosure. Terms clause 8.4 says that where a party is compelled by law to disclose the other's Confidential Information, it will, as far as the law permits, notify the other party in writing promptly and before disclosing, so that party can seek a protective order or other remedy. It will also give reasonable help in opposing the disclosure, and it limits disclosure to the portion legally required.
The obligation is mutual and sits in the published customer agreement, not a policy page. No transparency report was located on 1 September 2026.
The confidentiality clause requires prompt written notice when disclosure is compelled by a court, subpoena or government authority, to the extent the law permits notice. The receiving party must cooperate with efforts to obtain protection and disclose only the minimum information required by the legal process. Customer data and privileged material fall within the agreement's confidentiality framework. No accompanying transparency report describes the number or outcome of requests. This notice obligation is separate from the security incident process for unauthorized access.
Primary Law Corpus Provenance
Where does the law in this product come from, and does the vendor have the right to use it?
The product does not retrieve primary law, so there is no legal corpus to source. Lia operates on the firm's own case record and on what a client says by text or call about appointments, providers and symptoms. No public material identifies any statutory or case law source, license basis or update cadence, checked across the home page, product page, privacy policy and Terms and Conditions on 1 September 2026. Recorded as not addressed because the question does not arise for this product class rather than because the vendor declined to answer it.
The agent works with information supplied by a law firm and its clients, including session answers, attachments and recorded speech. Discovery support gathers facts for the attorney rather than drawing legal conclusions from a named research collection. No primary law source, licensing arrangement or update frequency is identified behind generated progress reports and conversation replies. Uploaded records remain customer material, even if they contain legal text.
The ability to classify files or organize responses does not identify an external collection of judicial or statutory authority.
Good Law Verification
Does the product tell you when the authority it just cited has been overruled?
No citator, and none would apply. Lia produces client communications and structured treatment records, and is expressly configured not to give legal advice or opinions, so there is no authority whose subsequent history could be checked. Nothing on any page read on 1 September 2026 addresses legal authority.
Trailmate organizes client facts and files within an assigned communication workflow. Its transcripts and summaries can be reviewed by the firm, and the agreement requires validation before reliance on generated material. No named citator or service for subsequent judicial treatment is part of that workflow. A reminder deadline and a completed document request concern case administration rather than the currency of legal authority.
The product does not describe a good law indicator when a participant mentions a case or when legal language appears in an uploaded file.
Refusal and Uncertainty Behavior
What does the product do when the answer is not in the corpus?
The vendor documents a category of question the system will not answer and describes what happens instead. The FAQ states the guardrail is built in, that Lia never provides legal advice or legal opinions, and that any question touching legal strategy is escalated directly to the assigned case manager. The same section states that a client mentioning they want to stop treatment or flagging a new provider also triggers immediate escalation.
That is an abstention path with a defined handoff rather than a confidence score. Recorded as documented rather than demonstrable because no published evaluation or observable test of the behavior was located on 1 September 2026.
The firm can define escalation points and configure a pause for sensitive information before engagement. Priority updates also identify work needing attorney or staff attention. Those controls route situations according to the assignment and professional sensitivity, but they do not specify what the agent says when it lacks factual support for an answer. No confidence measure, unsupported answer example or evaluated abstention behavior is provided. The configurable pause concerns sensitive conversations rather than a demonstrated response to missing information.
Fabricated Citation Record
Does a public court record exist addressing fabricated or hallucinated legal citations in output from this product?
No court order, opinion or disciplinary record naming this product has been located. The AI Hallucination Cases database maintained by Damien Charlotin was searched on 1 September 2026 on both the company name and the agent name Lia, alongside general sanctions coverage, and nothing naming the product was found. This is a statement about the public record rather than a finding about the product. Lia produces client correspondence and treatment records rather than citations to legal authority.
Damien Charlotin's AI Hallucination Cases database has no entry matching Trailmate concerning fabricated legal citations as of the stated date. Trailmate primarily conducts client conversations and turns collected information into structured updates for law firm review. Its terms warn that generated summaries and transcripts may contain errors and leave verification with the subscriber. No matching court finding was identified for this product.
That absence concerns fabricated citations specifically and is not a finding about the quality of its intake decisions, transcriptions or extracted case facts.
Bar Guidance Alignment
Has the vendor engaged in public with the ethics opinions its buyers are bound by?
No located public material engages with bar or ethics guidance. ABA Formal Opinion 512 is not named and no state bar opinion appears across the home page, the AI case agents page, the privacy policy or the Terms and Conditions, checked 1 September 2026. The vendor does name two statutory regimes it operates under, HIPAA and TCPA, and Terms clause 3.1(c) places responsibility on the firm to obtain client approvals for automated text messages, but neither is engagement with the professional responsibility guidance its buyers are bound by.
The agreement expressly assigns the subscriber responsibility for professional rules, bar requirements, competent representation and client confidentiality. It also requires consent appropriate to AI use in legal workflows. Product guidance gives an example introductory disclosure and describes limiting data collection before an engagement. These materials engage with professional responsibility in general terms without naming an ethics opinion or mapping controls to particular jurisdictions.
The firm's task settings and the vendor's software disclaimer therefore do not constitute a jurisdiction specific professional compliance guide.
Billing and Fee Posture
Does the vendor address what happens to the bill when the work takes an hour instead of six?
Time savings are published prominently and the billing consequence is not addressed. The product interface shows hours saved per week, and the site claims Lia manages 2,000 clients at once. The FAQ raises the question directly, asking how pricing works and whether the cost can be passed through to clients, and answers by referring the buyer to a sales rep and noting how other firms handle it. That is passing a technology cost through, not the treatment of AI assisted work on a bill. No audit record of which client messages were AI generated is described. Checked 1 September 2026.
Progress reports identify attempted work, completed tasks and collected information, while session transcripts and associated files can be monitored and exported. Those records preserve the activity of an AI agent on a matter, including outreach that might otherwise require staff time. The product also promotes reduced administrative work. It does not explain how firms should charge clients for that work or disclose the assistance in an invoice.
An operational task record is available, but it is not paired with a policy for billing hours or passing through the subscription cost.
Outside Counsel Guideline Readiness
Can a firm get this vendor through a client’s AI clause without a bespoke negotiation?
The material exists behind a request. The compliance section states that full documentation is available on request and that the SOC 2 Type II report is available on request, and no subprocessor list is published anywhere on the site. Terms clause 1.6 reserves the right to engage subcontractors to perform the Services in Provider's discretion without naming them or notifying the customer, and no model provider is identified, so a firm could not evidence whose systems see client content without contacting Hona. Checked 1 September 2026.
The agreement contains confidentiality, training and disclosure obligations, and product guidance offers a brief introduction for clients interacting with the AI assistant. No named model provider register or current list of product subprocessors accompanies those materials. The introductory message explains the assistant's role rather than identifying the organizations that process client content. A firm therefore has some consent language but no complete vendor package for satisfying a client's AI clause. No dedicated outside counsel disclosure pack or procedure for obtaining one is identified.
Court Disclosure Support
If a judge’s standing order requires an AI disclosure, can the product produce one?
Parts of a record exist, without an export designed for disclosure. The product page says every conversation Lia has is visible to the team, that escalations carry context including what happened, what she tried and what the client said, and that calls placed from the dashboard are recorded, transcribed and fed back. So what the AI sent and what the client replied is captured and attributable. No model is identified against any message, no human verification step is recorded, and no export framed for a disclosure duty is described.
Checked 1 September 2026. The product writes client correspondence rather than court filings, so the question rarely reaches it directly.
A session retains the conversation, attachments, generated information and task progress, and the firm can export the transcript and collected files. Those records could show the facts supplied during intake or discovery preparation. They do not combine an output with its model version, retrieved materials and a recorded human verification decision. Legal advice and validation remain with the firm, without a specific court disclosure certificate generated by the product. The available history supports reconstruction of part of the work rather than a complete certification export.
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 Hona if
- The main workload is keeping personal injury clients engaged through medical treatment.
- Missed appointments and treatment gaps should trigger configurable alerts.
- Case managers need to enter an active AI conversation when an exception needs attention.
Choose Trailmate if
- Client assignments vary across intake, document gathering, discovery and matter closing.
- The firm needs a common conversation workflow across several practice areas.
- Staff want to describe the task and reuse its process at selected case milestones.
In summary
Hona
Lia handles appointment coordination and treatment updates for personal injury firms. It collects symptoms and provider details, flags missed appointments or gaps, and lets staff enter a conversation when needed. Hona connects with named case management platforms.
Trailmate
Trailmate turns a firm instruction into a client conversation and a sequence of follow ups. Its workflows span multiple practice types, with configurable deadlines, data boundaries and escalation points. Client participation begins through the website or outreach links.
Questions buyers ask
Does Hona contact clients proactively?
Yes. Lia coordinates appointments, sends reminders and follows up after visits to collect treatment details. Staff can intervene, and configured rules control escalation. From the AI Legal Index, based on each vendor's own published materials as of October 16, 2026. No vendor pays for placement.
Is Trailmate restricted to personal injury?
No. It also describes workflows for immigration, employment, family law, criminal defense, trusts and estates and other consumer practices. From the AI Legal Index, based on each vendor's own published materials as of October 16, 2026. No vendor pays for placement.
How do people supervise the agents?
Hona surfaces treatment exceptions and allows staff to step into a conversation. Trailmate reports assignment progress and lets firms define deadlines, scope and escalation points. From the AI Legal Index, based on each vendor's own published materials as of October 16, 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.
Hona's Lia is the treatment focused reference here. Trailmate is operated by Trailmate, Inc.; this page compares software workflows rather than legal services.