JusticeText vs TrialKit: how they compare in 2026
JusticeText and TrialKit both sell AI discovery review to public defenders and private criminal defense firms. Both turn body camera video, audio and documents into something a lawyer can search and question, with answers tied back to the source. The sharpest difference is how each treats the case file. TrialKit states that privileged communications and work product stay scoped to the office, that each case runs in its own environment with its own indexes and AI outputs, and that customer material does not train general models. JusticeText's terms carry no confidentiality clause for customer content and let it tune its own models on that content. JusticeText's lead lies elsewhere: its terms bar relying on machine transcripts directly as evidence, it names AWS in the United States as its host, and it imports from Clio and cloud drives. TrialKit also states SOC 2 Type II and ISO 27001. It identifies the people who appear in video footage and shows when each one appears.
Which one fits
Choose JusticeText if
- Your clients' recordings are in many languages or on jail call systems. JusticeText transcribes and translates in more than 100 languages, reads GTL jail calls, For The Record courtroom audio and WatchGuard video, and syncs multiple camera angles by capture time.
- You want the vendor's terms to say plainly what its transcripts are for. JusticeText's terms call machine transcripts review aids, require the user to check every one, and bar relying on them directly as evidence, while the transcript itself flags passages of low confidence.
- Your evidence already lives in Clio or a cloud drive. JusticeText imports from Clio, Google Drive, Dropbox, OneDrive and Box, and stores data encrypted on AWS servers in the United States, with a deletion right completed in most cases within 30 days.
Choose TrialKit if
- You need each case walled off, down to the AI's own indexes. TrialKit states that every case runs in its own dedicated environment, that search indexes, embeddings and AI outputs are generated and stored per case and never queried across organizations, and that access is role based within the office.
- Your discovery includes phone extractions, handwritten notes and hours of surveillance video. TrialKit takes them in one upload, categorizes and names each file, identifies the people who appear in footage and when, and answers questions with citations to the exact file, page or timestamp.
- Your office's IT review needs certifications. TrialKit states SOC 2 Type II and ISO 27001 certification, 256 bit encryption at rest, audit logs and regular penetration testing, and shows the Los Angeles County Public Defender's Office and Reed Smith among its customers.
At a glance
All 15 axes, side by side
The same grid applied to every vendor in the index, graded from public sources. Hover a grade to see what the letter means on that axis.
AI Centrality
How much of the product is actually AI. Whether the machine learning is the mechanism the buyer is paying for or a feature layered onto conventional software, and whether the vendor is specific about which is which.
The models are the engine of the core capability, on an evidence platform with substantial functions that do not use AI. Speech recognition produces the searchable transcripts, synced to the media, in more than 100 languages that the product leads with, and the MirandaAI assistant, summaries, key moments and PDF analysis are model work. Around them sits conventional evidence tooling that works without models: conversion of proprietary formats such as GTL jail calls and WatchGuard video, syncing of several camera angles by capture time, and clipping, redaction and captioning for courtroom exhibits.
The models are the product. Ingestion, categorization and naming of files, extraction from phone dumps and handwritten notes, identification of people in video, mapping of people, places and entities, timelines and cited answers to questions are all model work. Without them, what remains is file storage.
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.
Output is grounded in the source media, and a failure condition is named with a visible behavior, short of any measurement. Transcripts are synced in time to the original audio and video, so any passage can be checked against the recording. The product FAQ names the conditions under which transcription degrades, overlapping voices, static and background noise, and says the transcript clearly identifies areas of lower confidence so reviewers focus there. No accuracy figure is published, nothing describes how MirandaAI answers are grounded or cited, and the terms disclaim any warranty of transcript accuracy.
Answers are tied to the record at a fine grain. TrialKit says it answers questions in seconds, citing every claim to the exact file, page or timestamp, with every answer traced back to its source. No accuracy figures, test set or error rate are published for answers, extraction or the people it identifies in footage.
Autonomy and Oversight Model
What the system decides on its own, what a lawyer must approve, and whether the vendor documents where the review point sits. A tool that drafts under review and a tool that files without one are different products and different risks.
A named output carries a categorical limit on use, with a review surface built in. The Terms of Service state that machine generated transcriptions are tools to aid review, not authoritative records of events, and that every transcription must be checked by the user. Users agree not to rely directly on system transcriptions as evidence in a court, arbitration, mediation or other formal proceeding or settlement negotiation. The product supports that check by flagging passages with low confidence and syncing text to the original media, and exhibits are clipped from the original video rather than the transcript. The limit is specific to transcriptions: MirandaAI's timelines and inconsistency findings carry no equivalent stated limit.
Oversight rests on traceability and roles. Cited sources let a lawyer check each answer, and workspaces have roles and tagging. Nothing describes a review step, a limit on relying on an answer or an identification without checking, or what happens when the system 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 deployments with figures, short of any method. The vendor publishes case studies and rollout stories naming the Kentucky Department of Public Advocacy, the Santa Cruz County, Lane County, Travis County, Lehigh County and Siskiyou County defenders, Kansas Holistic Defenders, and private firms including Rossen Law and Nicholson Goetz & Otis. It quotes chief defenders in Harris County and Montana by name. Rossen Law says the product cut video review and trial preparation time by at least two thirds. No figure states how it was measured, over what caseload or period.
Named customers are shown without measured results. The home page shows logos for the Los Angeles County Public Defender's Office, the Capital Defense Project of Southeast Louisiana and Reed Smith, among other defense and family law firms. It quotes named attorneys such as Zach Boyd of Tyroch Boyd PLLC on saving weeks of preparation and Patrick Beasley of Beasley Law Group on identifying a discovery violation. No customer outcome is measured, dated or explained.
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.
Confidentiality is asserted in general terms, while the agreement permits model tuning from customer content. The product page says data remains confidential, the customer controls access and deletion, content is encrypted on U.S. servers, and evidence is never used to train third party machine learning models. The Terms of Service contain no confidentiality clause protecting customer content. The content license is limited to providing the service, but a separate Machine Learning clause authorizes calibrating data models and improving algorithms while processing user content. The privacy policy permits sharing deidentified aggregate data with research partners and funders. Nothing addresses privilege or defense work product, which is what public defenders' notes, tags and clips are.
Specific commitments sit on a policy page. The Security page says case files, work product and privileged communications are scoped to the office or firm and never pooled across customers or shared, and that case data is isolated per case with access by role. It also says customer material does not train general AI models, reinforced through contractual terms. The contract itself is not published, and no position is stated on which outside AI providers, if any, see case material.
UPL and Professional Responsibility Posture
Whether the vendor is clear that it supplies a tool rather than legal advice, who its audience is, and how it addresses unauthorized practice of law, competence and supervision duties, and jurisdiction limits. ABA Formal Opinion 512 is the reference point. Where the advice line is not the duty a product raises, the axis is read through the nearest professional duty it does raise: judicial conduct rules and the reviewing duty for products sold only to courts, and the duty to bill for time actually spent for products that draft time entries.
A clear position on what the output is and is not, with a defined audience, short of supervision duties. The terms state that transcriptions are review aids and not authoritative records, must be checked, and must not be relied on directly as evidence, and both the terms and privacy policy say JusticeText cannot answer requests for legal advice. The audience is attorneys, and the privacy policy collects each user's law licensing state. The terms disclaim any representation that the service is appropriate or lawful in a given location, including under court orders. Nothing addresses an attorney's competence or supervision duties when relying on MirandaAI analysis.
Nothing states what the product is and is not, that its answers and analysis are not legal advice, or how it supports a defense lawyer's supervision and competence duties. The only published terms govern use of the website, not the product.
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 on the vendor's own site. The product's Built for responsible review section covers data security only, and no responsible AI policy, testing regime or accountable owner is published. A submission by the vendor hosted by MIT Solve acknowledges that speech recognition engines are less accurate on African American Vernacular English. That is a disclosure of uneven output, but it is not on the vendor's own site and is not dated.
No AI governance framework, accountable owner, testing before release or bias disclosure is published. The home page says TrialKit identifies the people in video footage and shows when each one appears and for how long. Nothing published explains how that identification works, how accurate it is, whether accuracy differs across groups of people, or how its use is limited. The home, Security and About pages and the site's terms of use were checked.
AI Safety and Data Stewardship
Retention, deletion, access control, and what happens to prompts and documents after they are processed. Whether the vendor states its subprocessors and its incident practice, or leaves the buyer to assume.
The privacy policy states that information is stored encrypted on AWS servers in the United States and encrypted in transit, gives users a right to deletion completed in most cases within 30 days, and names categories of providers (hosting, transcoding, storage, databases, crash monitoring, payments) with AWS and Stripe named. The product page adds multi factor authentication and single sign on and says customers control access and deletion. Breach notification is promised only as required by law, the terms reserve retention and destruction under internal policies that are not published, and no full subprocessor list, including any AI or speech recognition provider, is given.
The Security page covers much of the ground: AES-256 encryption at rest and TLS 1.2 or higher in transit, access by role, audit logs of activity, tested backups, continuous monitoring, vulnerability scanning and regular penetration testing, and deletion or archiving of case data by authorized users. Retention has no stated period, and no subprocessor list or incident notification commitment is published.
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.
Liability is addressed only through limitations that disclaim the exposure the product creates. The terms provide the system as is, disclaim any warranty of transcript accuracy, disclaim liability for any legal action or court decision resulting from transcripts or machine learning output, cap liability at the amount paid or $100, and require users to indemnify JusticeText. No vendor indemnity or insurance position is published.
No customer agreement is published, so nothing states who bears the loss when an answer, extraction or identification is wrong. The site's terms of use cover only the website.
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 connections exist, short of documented depth. The product page shows imports from Google Drive, Dropbox, OneDrive, Box and the Clio practice management system, and the privacy policy describes the cloud storage integrations and Google API use. The platform also takes in the proprietary formats defense discovery arrives in, including GTL jail calls, For The Record courtroom recordings, AV Viewer and WatchGuard. What each connection moves, and whether results flow back to Clio or case management, is not documented.
Discovery arrives by upload. The home page describes the product as needing no setup or plugins, and no integration with a review platform, case management or document management system is named.
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.
The hosting model and region are stated, with no options and no tenancy detail. The privacy policy states that information is stored in encrypted form on servers in the United States operated by Amazon Web Services, and the product page repeats that content is stored on encrypted U.S. servers. No tenancy model, deployment option or separate processing location for speech recognition or AI providers is stated.
Isolation is described, and location is not. Cases run in dedicated environments of their own, and security testing covers all deployment environments. No hosting provider, region, residency option or processing location is published.
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.
No independent security attestation located. The vendor describes encryption, multi factor authentication and single sign on but names no audit or standard, which matters for a product holding criminal discovery.
Named attestations are claimed, without a report route. The Security page claims SOC 2 Type II and ISO 27001 certification and lists HIPAA as coming soon. No auditor, certification body, scope or date is given, no trust center is published, and no way to obtain the reports is described.
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 product page says evidence is never used to train third party machine learning models, and the terms say third party tools execute some functionality and may access user content. No speech recognition engine, language model or provider is named on the product page or in the terms, and no change notification is described.
No model, model provider, architecture or processing location is disclosed anywhere on the site, including for transcription or for identifying people in video.
Commercial Transparency
Whether a buyer can learn what this costs without entering a sales process: published rates, the unit being charged, what sits behind an enterprise tier, and what implementation adds.
The pricing structure is described without figures. The Terms of Service describe monthly and annual subscriptions that renew automatically, usage limits on processing time that can block further processing until the next term, free initial access for a period or amount of usage, and no refunds. No price, rate or plan is published; access starts through an interest form.
No price, tier or unit of charge is published; the home page offers custom pricing for litigation teams of every size. The home and demo pages were checked.
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 vendor serves public defender offices and private criminal defense attorneys, and covers body camera footage, interrogations, jail calls, dashcam, courtroom recordings and written discovery, with transcription and translation in more than 100 languages and support for named proprietary formats. It does not state which jurisdictions it serves, which evidence types it handles poorly beyond noisy audio, or whether it suits civil or prosecution use.
The core buyers are clear: public defender offices, capital defense projects and private criminal defense firms, with civil litigators and family law firms also shown, from solo practitioners to national firms. The evidence types covered are set out in detail, from body camera footage and phone extractions to handwritten notes. What the product does not handle, by file type, volume or matter type, is not stated.
The 12 legal signals, side by side
Recorded rather than graded. These are the questions a practitioner has to answer before a tool touches a client matter, and the answers are taken from public material only.
Client Data in Training
Can material a lawyer puts into this product be used to train a model?
The published Terms of Service reserve a right to tune the vendor's own models using customer content, with no opt out located. The Machine Learning clause authorizes JusticeText to use machine learning to generate statistics, calibrate data models and improve algorithms in the course of processing uploaded content and usage data, and to test, tune, optimize, validate or otherwise enhance the models underlying the system.
The product page states that evidence is never used to train third party machine learning models; that statement is limited to third parties and does not address the vendor's own models.
The Security page states that TrialKit does not use customer inputs, outputs or uploaded documents to train or improve general AI models, and says contractual terms and internal controls reinforce this. Those terms are not published, and the commitment is framed around general models; indexes and embeddings for each case are described as stored per case.
Prompt and Output Retention
How long does the product keep what a lawyer typed, and can that be set to zero?
The customer controls deletion of uploaded evidence. The product page says customers control access and deletion, and the privacy policy gives a right to deletion completed in most cases within 30 days, subject to legal retention and to deidentified usage data the vendor may keep. The terms also reserve retention and destruction under internal policies that are not published.
The Security page says data is kept only as long as necessary to support active use, and that authorized users may delete case data or archive it under the company's retention policies, which are not published. The home page advertises zero data retention without saying what it applies to.
Ethical Walls and Matter Segregation
Does retrieval respect the firm’s ethical walls, or can the model read across them?
Sharing is controlled by the user, with no published detail on how separation is enforced. The privacy policy says data and work product are shared with other users only on request or according to account preferences, such as with another member of the same office. Nothing documents how cases, attorneys or codefendant matters are walled off within an office.
The Security page describes isolation by case: each case runs in its own environment that is never shared, reused or commingled, and metadata, search indexes, embeddings and AI outputs are generated and stored per case and never queried across organizations. Access is set by role, so attorneys, paralegals and staff see only what they need. The customer maintains the roles, and no inheritance from a document management system is described.
Third Party Request and Subpoena Notice
If someone subpoenas the vendor for a firm’s data, does the firm hear about it first?
Disclosure under legal compulsion is limited to what the law requires, and customer notice is not addressed. The privacy policy permits disclosure for compliance with law enforcement and court orders only to the extent required, and the terms allow disclosure to law enforcement on breach. Neither commits to telling the customer before or after disclosure, which matters for a product holding defense discovery and work product.
No product privacy policy or customer agreement is published, and nothing addresses how requests from authorities or other third parties for case data would be handled. The home and Security pages and the site's terms were checked on 22 September 2026.
Primary Law Corpus Provenance
Where does the law in this product come from, and does the vendor have the right to use it?
Licensing of legal content is a marginal question for a tool built on each customer's discovery. The AI works over the customer's own discovery rather than a legal research corpus, and no external legal source is described.
The product works on the discovery each customer uploads, and no body of primary law behind its output is identified. Checked on 22 September 2026.
Good Law Verification
Does the product tell you when the authority it just cited has been overruled?
Good law status has nothing to attach to here. The product does not cite legal authority, so no check of subsequent history arises and none is described.
The product cites case evidence rather than legal authority, and nothing addresses checking authority for later history, as of 22 September 2026.
Refusal and Uncertainty Behavior
What does the product do when the answer is not in the corpus?
The transcript marks passages with low confidence, with no documented abstention path. The product FAQ says that for poor audio, overlapping voices and background noise the AI transcript clearly identifies areas of lower confidence so reviewers can focus there. Nothing describes what MirandaAI does when the evidence does not answer a question.
No confidence indicator is described for answers or for identifying people in video. Nor is any path in which the system declines when the record does not support an answer, as of 22 September 2026.
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 kept by Damien Charlotin holds no entry naming JusticeText as the tool used. JusticeText transcribes and summarizes evidence recordings for criminal defense teams.
The AI Hallucination Cases database kept by Damien Charlotin holds no entry naming TrialKit as the tool used.
Bar Guidance Alignment
Has the vendor engaged in public with the ethics opinions its buyers are bound by?
Lawyers using the product get no pointer to the ethics rules that govern that use. No bar ethics opinion or court rule on AI is named or engaged in connection with the product.
No material refers to lawyers' professional or ethical obligations, and no ethics opinion is named, on the home, Security or About pages or in the site's terms as of 22 September 2026.
Billing and Fee Posture
Does the vendor address what happens to the bill when the work takes an hour instead of six?
The product is bought mainly by public defender offices, where no client is billed for the work. Time savings are claimed as capacity for the office's own caseload. Private criminal defense firms also use it, where a fee relationship would arise, and nothing addresses billing or disclosure for that use.
Attorneys are quoted on time saved, and a journalist on work that took 30 hours taking 30 seconds. Nothing addresses how private defense firms bill or disclose work done with the product; public defender offices bill no client.
Outside Counsel Guideline Readiness
Can a firm get this vendor through a client’s AI clause without a bespoke negotiation?
Infrastructure providers are named, and the AI providers a client would ask about are not. The privacy policy names AWS and Stripe and lists other provider categories, but no speech recognition or AI provider is named and no data processing agreement is published.
No subprocessor or model provider list is published, as of 22 September 2026.
Court Disclosure Support
If a judge’s standing order requires an AI disclosure, can the product produce one?
Transcripts are synced in time to the original media with passages of low confidence marked, and courtroom exhibits are clipped from the original video with captions. The terms bar relying on the transcript itself as evidence. Nothing records which model produced a transcript, summary or MirandaAI answer, or who verified it.
Answers cite their sources at file, page or timestamp level, and the Security page describes logs with timestamps for user and system activity. No export of a record showing the model used, the sources and the human check is described, and no disclosure guidance is published.
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
- Primary Law Corpus Provenance
- Good Law Verification
- Bar Guidance Alignment
- Outside Counsel Guideline Readiness
In summary
JusticeText
JusticeText is built mainly for public defender offices and also serves private defense firms, from JusticeText Inc. in Irvine, California. Its speech recognition turns body camera, interrogation, dashcam, courtroom and jail call recordings into transcripts synced to the media in more than 100 languages. Its MirandaAI assistant summarizes, finds key moments and checks documents against video. According to the AI Legal Index, JusticeText's terms treat transcripts as review aids that must not be relied on directly as evidence, and it stores data on AWS servers in the United States. Its terms also allow it to tune its own models on customer content. Access starts through an interest form.
TrialKit
TrialKit, from TrialKit Inc. of Clifton, New Jersey, is an AI discovery platform built by attorneys for criminal defense and litigation teams. It takes documents, body camera and surveillance video, audio, phone extractions and handwritten notes in one upload, then names and extracts them and identifies the people in footage. It maps people, places and dates into a timeline and answers questions with citations to the exact file, page or timestamp. According to the AI Legal Index, each TrialKit case runs in its own isolated environment, and the vendor states SOC 2 Type II and ISO 27001. It offers custom pricing for litigation teams of every size.
Questions buyers ask
JusticeText vs TrialKit: which is better for public defenders?
Both turn body camera video, audio and documents into something a defense lawyer can search and question. JusticeText's terms set how its transcripts may be used, and it names where data is stored and what it imports from. TrialKit commits to isolating each case and keeping customer material out of general model training, and states security certifications. Offices weighing confidentiality of the case file first can start with TrialKit's case isolation. From the AI Legal Index, based on each vendor's own published materials as of September 25, 2026. No vendor pays for placement.
Does TrialKit keep one case's data apart from another's?
Yes, by its own description. TrialKit's security page says every case runs in its own dedicated environment that is never shared or reused, that metadata, search indexes, embeddings and AI outputs are stored per case and never queried across organizations, and that case files and privileged communications are never pooled across customers. Access is role based. From the AI Legal Index, based on each vendor's own published materials as of September 25, 2026. No vendor pays for placement.
Does TrialKit identify people in body camera video?
TrialKit's home page says it identifies the people who appear in video footage and shows when each one appears and for how long. JusticeText's MirandaAI finds key moments in footage and checks documents against video. From the AI Legal Index, based on each vendor's own published materials as of September 25, 2026. No vendor pays for placement.
Which one states security certifications?
TrialKit's security page states SOC 2 Type II and ISO 27001 certification, with HIPAA listed as coming soon, alongside 256 bit encryption at rest, audit logs and regular penetration testing. JusticeText describes encryption, multi factor authentication and single sign on, with hosting on AWS in the United States. From the AI Legal Index, based on each vendor's own published materials as of September 25, 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.
Two points to weigh. JusticeText's terms permit it to calibrate and tune its own models on customer content, a published term. Its product page's no training statement covers third party models only. TrialKit identifies people in video footage and publishes nothing on how that identification works, how accurate it is, or whether accuracy differs across groups of people. Its no training commitment rests on contract terms that are not published. Neither vendor reviewed this page.