Gideon

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.

A conversational intake and CRM product with a genuine analytics model inside it, where the workflow rather than the model is what the buyer purchases. The chatbot layer is described as easy to build and driven by firm specified qualification criteria, and independent comparison material characterises the product as a chatbot bundled with an intake workflow tool where form data flows into Gideon's own CRM. That is configured conversational automation plus a CRM, and both function on rules rather than on generative capability. The genuinely model driven component is the predictive analytics engine, stated to learn a firm's preferences for particular case types and predict likely case outcomes, which is a real machine learning claim and is the vendor's oldest differentiator, present in its positioning since 2017. Held at C because removing the models leaves a working intake chatbot, questionnaire builder, scheduler, document generator and CRM, which is most of what a firm buys. Same placement as LawDroid and for the same reason: this category's products are automation platforms with models added rather than models with interfaces.

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.

Nothing published, and the axis applies in an unusual form on this record. The product generates no legal citations and performs no legal research, so the classic grounding question does not arise. What does arise is the accuracy of the predictive analytics engine, which is stated to predict likely case outcomes so firms can direct resources toward higher value matters, and that is a quantitative prediction about a real world result that is testable in principle and published nowhere. No accuracy figure, no calibration statement, no error rate on qualification, no evaluation and no methodology were located for either the prediction engine or the chatbot's answers to initial enquiries. A prediction engine whose accuracy is unstated is being relied on to decide which prospective clients a firm pursues, and neither the firm nor the consumer can assess it. Checked the home page, the blog material, the Clio directory listing, the CodeX profile and independent comparison material on 29 Aug 2026.

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.

Both directions are stated plainly and they sit in tension, which the note records rather than resolves. The vendor states that chatbots qualify leads and route them to the right attorney or staff person automatically, direct from the website, and expressly without any human intervention, which is an unambiguous autonomy claim about a system conversing with the public. It separately states that when a lead is qualified the relevant attorney or staff member is alerted immediately and can jump in to convert the lead in real time over messaging, which is a human takeover mechanism of the same shape credited on LawDroid. So a human can enter the conversation and the qualification and routing decision has already been made without them. Held at C rather than B because the takeover is described as a sales opportunity rather than as an oversight control, no trigger or threshold is published, nothing states what the bot may say unattended before a human arrives, and the decision that carries consequence for the consumer, being whether they are qualified at all, is the one explicitly performed without human intervention.

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.

Institutional recognition and funding are checkable, and customer evidence is absent. Verifiable without the vendor: a Stanford Law School CodeX startup profile featuring a named cofounder, Elan Fields, which is an academic legal technology programme rather than a marketing placement; a published integration listing in the Clio app directory, which required a partner relationship with the largest practice management vendor in the market; independent secondary market data reporting $11.5m total funding with a seed round in March 2026; and sustained independent comparison coverage assessing the product against named competitors. Longevity is real, with positioning documented from 2017. Against that: no law firm customer is named anywhere in located material, no case study, no usage figure, and no outcome claim with a figure despite the product being built around measurable conversion metrics, which is a conspicuous omission for an intake platform whose entire value proposition is lead conversion. Held at C on that split.

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.

Nothing located. No confidentiality statement, no encryption reference, no data handling description and no treatment of privilege or prospective client confidences was found in any surface read. The gap is the category's central one and is sharper here than on LawDroid because this product retains the intake data in its own CRM rather than passing it through: consumers describe their legal problems to a chatbot on a firm's website, that account is captured, analysed by a prediction engine and stored in the vendor's system, and nothing published addresses the status of those communications, whether they are treated as confidential, or what happens to them when the firm declines the matter. Checked the home page, the blog material, the Clio directory listing, the CodeX profile and the site navigation on 29 Aug 2026.

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.

Not located. The chatbot is stated to answer initial enquiries from consumers and to qualify them against case type criteria without human intervention, which places an automated system in the first conversation a person has about their legal problem. Answering an initial enquiry is where the line between legal information and legal advice is thinnest, and telling a consumer they do or do not have a qualifying matter is a determination they will act on. Nothing published states that the chatbot does not give legal advice, addresses what it may and may not say, describes any disclosure to the consumer that they are talking to a machine, or engages any bar guidance on automated client communication. Distinguished from LawDroid, which markets legal guidance to the public as a product; here the exposure arises from qualification and enquiry handling rather than from advice being sold, and it is unaddressed either way. Checked the home page, the blog material, the Clio directory listing and the CodeX profile on 29 Aug 2026.

AI Governance and Bias Disclosure

Published governance over model behaviour: who owns it inside the vendor, what is tested before release, and what is disclosed about disparate output across matter types, parties, or populations.

Nothing published, and this record carries the most concrete untested bias risk in the category. The predictive analytics engine is stated to learn a firm's preferences for case types and predict likely case outcomes so firms can deploy resources to optimum value, which is a model scoring prospective clients on expected value and directing firm attention accordingly. A system trained on which past matters a firm accepted and how they resolved will reproduce whatever patterns are in that history, and the people affected are consumers seeking legal help who are never told a prediction was made about them, cannot see it, and have no route to contest it. The vendor's stated mission is making legal help accessible to everyone, which makes the absence of any fairness evaluation more pointed rather than less. No AI policy, model card, bias testing, evaluation, monitoring, governance body, ISO 42001 or EU AI Act positioning was located.

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.

No stewardship position located. Nothing states whether intake conversations, captured matter details, generated documents or outcome data are used to train or improve models, no retention period is published, and no deletion right is described. The architecture makes the question weightier than on peers that pass data straight through: independent comparison material describes the CRM as part of the same product, so consumer intake data is retained by the vendor rather than only transiting, and the prediction engine is stated to learn from case outcomes, which means outcome data from firms feeds model improvement without any published statement about whose data, on what basis, or whether learning is scoped to the individual firm or pooled across customers. That last question is the one a firm should ask and it is unanswered. Checked the home page, the blog, the Clio listing, the CodeX profile and independent material on 29 Aug 2026.

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.

No published position located. Nothing was found on liability for AI output, warranty, service levels or remedy. Two exposures are specific to this product. A chatbot that wrongly disqualifies a consumer ends the conversation, and unlike a defective document or a bad citation nobody discovers the error, since the consumer simply goes elsewhere or gives up and the firm never learns what it turned away, which makes it the least visible failure mode graded on this index. A missed or mishandled enquiry can also carry limitation period consequences where a matter is time sensitive. Nothing published addresses either, and no service level commitment was located for a system described as an always open front door. Checked the home page, the blog material, the Clio directory listing and the site navigation on 29 Aug 2026.

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.

A named practice management integration published on the partner's own directory, plus an architecture that removes the integration problem for part of the workflow. Clio integration is listed in the Clio app directory, which is verification by the counterparty rather than a vendor claim, and Clio is the practice management system this buyer segment most commonly runs. The architectural point is credited because independent comparison material makes it precisely: because the chatbot and the intake CRM are the same product, no middleware is needed to connect them, which the reviewer identifies as a no handoff workflow and a genuine strength. The same material identifies the trade off honestly, noting the CRM half is redundant for a firm already running Lawmatics, Clio Grow or Filevine, and that channel coverage is website focused with limited direct messaging and WhatsApp support. Held at B because only one practice management vendor is named, no API documentation was located, and the vendor's stated reach across directory profiles and social media is not matched by named platform integrations.

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.

Nothing located. No hosting provider is named, no region or data residency commitment is published, and no deployment options are described. The product is a website embedded chatbot with a hosted CRM behind it, so consumer intake data plainly resides with the vendor, and nothing states where. Residency is a live question because the chatbot collects personal information from members of the public who may be in any jurisdiction, and state privacy regimes attach to that collection regardless of where the firm sits. Checked the home page, the blog material, the Clio directory listing, the CodeX profile and the site navigation on 29 Aug 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.

No certification, attestation, trust centre or security page was located. No SOC 2 of either type, no ISO 27001, no named auditor, no penetration testing partner and no encryption statement were found across any surface read. Under the three tier test the artifact is absent rather than gated. Two records in legal-intake-and-client-development now sit at D on this axis, this one and LawDroid, and both are the small vendors in the category, which is worth watching as the remaining five are built rather than concluding anything from two. The gap is material for a product holding consumer personal information collected through a law firm's website, where the firm carries the obligation and cannot evidence the vendor's controls to anyone. Checked the home page, the blog, the Clio directory listing, the CodeX profile and the site navigation on 29 Aug 2026.

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 located. No foundation model provider, model family or version is named, no distinction is drawn between proprietary and third party models, and no subprocessor list was found. The predictive analytics engine is described as the vendor's own capability, which is an ownership statement rather than a supply chain disclosure, and independent classification of the company references natural language and conversational AI without identifying what powers it. For a platform processing consumer personal information on a law firm's behalf, the identity of any third party processing that data is a question the firm's own obligations require it to answer and it cannot be answered from public material. Checked the home page, the blog, the Clio directory listing, the CodeX profile and independent material on 29 Aug 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.

No pricing published at any level. No price, no range, no tier structure and no unit of charge, and no indication of whether the platform prices per seat, per conversation, per captured lead or per firm, which matters because those scale very differently for an intake product and determine whether cost rises with marketing spend. Independent comparison material positions the product as suited to mid size firms wanting a chatbot and intake CRM in one, without publishing a figure. Notable within this category: LawDroid publishes complete pricing across three products and reaches A on this axis, and this record publishes nothing, so the category is not uniform and the contrast is between vendors rather than inherent to intake products. Checked the home page, the pricing navigation, the Clio directory listing and independent comparison material on 29 Aug 2026.

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.

Channel coverage is characterised and practice coverage is not. The vendor states reach across the firm website, directory profiles and social media, which is a real channel claim for an intake product since prospective clients arrive from all three, and independent comparison material qualifies it usefully by noting the product is website focused with limited direct messaging and WhatsApp coverage, so the stated reach exceeds the verified reach. Firm size targeting is identified by independent review as mid size firms. What is absent: no practice areas are named, no jurisdictions are stated, and because qualification criteria are configured by the firm the substantive coverage is whatever the customer builds, which the vendor does not say. The predictive analytics engine is stated to learn firm preferences by case type, which implies practice area handling without enumerating any. Held at C on that basis.