L

LegalClerk.ai

The models drive what a firm buys, on top of a working telephony platform. The paid plan is sold as unlimited AI calls, the AI answers and conducts every call, and no human receptionist is offered on any tier except an optional live agent escalation at enterprise level, so the purchase is the automation. The product beneath, though, is a phone system that works without any model. The published agreement describes an online phone and call handling service, with number allocation, number porting, per minute rates, answer supervision and codec requirements, and never mentions artificial intelligence.

LegalClerk also ships a Chrome extension for parallel dialing and click to call, which is manual outbound dialing rather than an AI feature. Its privacy notice describes the company as an AI dialing solution for businesses globally rather than a legal intake product. Without the models, a telephony service remains.

Accuracy is asserted repeatedly and never measured. The claims go directly to output quality: the system never misses a required field or form detail, follows the intake script every time with zero room for human error, forgets no fields, mishears no details and eliminates intake errors. None comes with a figure, test set, evaluation, transcription error rate or statement of what happens when the system mishears a caller.

Nothing addresses hallucination, and there is no accuracy or benchmark page. Grounding to primary authority and citation status do not apply, since the system asks scripted questions and records answers rather than retrieving or citing law. The claims matter because the same pages describe the system detecting accident types, timestamping incidents against limitation periods and classifying injury severity, all from a caller speaking on a phone line.

The system runs a whole client conversation unattended, and the controls around it are real and published. A firm sets when the AI answers, writes the script it follows, defines the screening criteria, and chooses which calls are transferred or escalated as urgent. Serious matters are flagged for a priority attorney callback, and the enterprise tier adds optional escalation to a live agent. The review surface is substantial: every call is recorded and transcribed, a written summary and structured lead profile are delivered, and call analytics and lead reports are available.

No lawyer is involved while the system speaks to a prospective client, and nothing is published about what happens when it is wrong: no error handling, correction path, abstention behavior, or plan for a caller it cannot understand. LegalClerk's FAQ states that most callers cannot tell it is AI and that this is by design. Hiding the automation from the person on the line is an oversight choice, published plainly.

No production evidence is published, and the social proof on LegalClerk's home page does not hold up. The page says the product is loved by lawyers across the country above three user avatars, but those images come from i.pravatar.cc, a placeholder avatar generator, requested by index number. They are not customers. The site also claims the product is trusted by top law firms and used by leading legal teams, with no firm, case study, named individual or dated deployment.

The published numbers are industry statistics and cost comparisons rather than outcomes. They include a claim that 39 percent of calls to law firms go unanswered, a comparison of a receptionist at 48,000 to 65,000 dollars a year with roughly 4,800 for the service, and a resulting 90 percent cost reduction. None of that shows what happened at a firm that deployed it.

Confidentiality is asserted in general terms in the marketing, and the agreement points the other way on the one clause that reaches call content. The product records and transcribes conversations with prospective clients, which is material a firm would treat as confidential from the first call. The marketing offers HIPAA grade encryption, secure storage, detailed access logging, secure storage plus redaction, and intake staff trained in client confidentiality.

The agreement does not carry any of that across. Its confidentiality section is a mutual nondisclosure clause covering proprietary information the parties disclose to each other, which is business information rather than caller material. Customer Data, expressly defined to include call recordings, voicemails and call metadata, is instead the subject of section 4.4: the customer grants a license to use, copy, transmit, distribute and store that data to the extent necessary or desirable to provide and improve the services.

No privilege or work product treatment appears anywhere, nothing describes tenancy or separation between firms, and there is no published data processing agreement, although the terms state one forms an inseparable part of the contract. Nothing addresses the prospective client relationship the product exists to create.

Nothing published addresses the advice line, on a product that speaks directly to members of the public about their legal problems. There is no statement that the company is not a law firm or does not give legal advice, no description of what the AI may say to a caller, no jurisdiction limit, and nothing on the firm's supervision or competence duties when an automated system handles first contact. The agreement is silent on the subject, and the pricing page's claim of HIPAA and ABA compliance names no authority and takes no position on advice.

The exposure comes from the vendor's own descriptions. The system timestamps an incident to flag possible statute of limitations issues, filters callers on whether another party was at fault, classifies injury severity to gauge case strength, and prompts callers about what they have told an insurer. By the vendor's account, most callers cannot tell they are speaking to software.

No governance position of any kind is published. There is no responsible AI page, principles statement, named owner, testing regime before release or management system. Nothing addresses uneven performance across the people this product serves. The system is sold for multilingual intake and immigration practice, and it is a speech system taking calls from distressed callers, so recognition performance across accents, dialects and languages is the obvious place for uneven output.

Nothing published tests or discloses it. The nearest statements concern encryption and access logging, which are security rather than governance. The only policy documents published are the terms and conditions and the privacy notice.

Access and encryption are addressed in marketing terms, and the privacy notice does not reach the recordings. Published: an AES-256 encryption badge, encryption in transit and at rest, access restricted to authorized personnel, detailed access logging, regular security audits, and secure storage with redaction as a plan feature. What governs the material this product creates is absent. No retention period is stated for call recordings, transcripts, summaries or lead records, no deletion or return of data commitment covers them, no subprocessor list exists, and no breach notification commitment is published.

The privacy notice, last updated 15 September 2025, describes the company as an AI dialing solution and lists the personal data it collects as contact, account, usage and technical data. Most of it concerns cookies, social plugins, Google OAuth tokens and a Chrome extension addendum. Call content is not mentioned, so nothing published governs the recordings of prospective clients the service exists to make. Its deletion and rights routes point to support addresses at a different company's domain.

LegalClerk publishes how loss is allocated, and every part runs one way. Section 11 caps cumulative liability at the fees paid in the preceding twelve months, or one hundred euros during a free trial, and excludes consequential, indirect, incidental, reputational and punitive damages including loss of data and lost business opportunity, with any claim required to be served by registered post within one year. Section 9.3 provides the service as is and as available and disclaims all implied warranties including fitness for purpose, warranting only a commercially reasonable level of care.

Section 10 runs the indemnity from the customer to LegalClerk, covering breach, violations of law and claims relating to Customer Data; there is no indemnity from LegalClerk anywhere. No warranty on output, service credit or insurance position is published, and nothing addresses a missed call, a misheard intake field or a caller mishandled by the system. The exposure is addressed and readable before signing, through a standard limitation clause that disclaims the risk the product creates.

Named connections into the systems a firm already runs, with what moves stated and no depth behind it. Clio, MyCase, Lawmatics, PracticePanther, Google Calendar and Calendly are named, against a claim of more than thirty five case management integrations, and each has its own page. What moves is stated at workflow level: new leads, booked appointments and intake summaries sync automatically, and intake data reaches the case management system as a lead profile.

Nothing an implementer would need is published: no field mapping, sync direction, trigger conditions, error or conflict handling, configuration guide or API documentation. Section 4.6 of the agreement disclaims integrations entirely. It says they are not LegalClerk's products, that their use is solely between the customer and the third party, and that the customer warrants it has every consent needed to move data in both directions. An integrations index is published.

Cloud delivery is implied throughout, and neither tenancy nor region is stated. Nothing says whether the platform is single or multi tenant, no dedicated or isolated option appears on any tier, and no region is named for storage or processing. The only geographic facts published are contractual, and they point somewhere unexpected for a product sold with HIPAA, PHIPA and PIPEDA badges. The agreement is governed by Slovak law with exclusive jurisdiction in the Slovak courts, notices go to a Bratislava address, fees may be in euros, the free trial liability floor is expressed in euros, and daily call thresholds run on Central European Time.

None of that says where customer data sits. Where speech processing happens is also unstated, since no model provider is identified. No published source says which country holds the recording of a call with a prospective client.

Four badges, with no scope, date, report or auditor. LegalClerk's home page shows compliance marks for HIPAA, PHIPA and PIPEDA and an AES-256 encryption badge, each an image with no statement of what was assessed, by whom, over what period or against which controls. The pricing page repeats the claim as HIPAA and ABA compliant. HIPAA, PHIPA and PIPEDA are statutory regimes a company attests to itself, not independent attestations, and no Business Associate Agreement is offered despite the HIPAA claim.

The site calls itself proven, audited and secure and refers to regular security audits, without naming an assessor or publishing a summary. There is no SOC 2 of either type, ISO certification, penetration test summary, trust center or portal, so there is nothing gated to request.

LegalClerk gestures at the technology and identifies nothing underneath it. The site refers to advanced natural language processing, human like voice technology, NLP tagging that classifies accident types, and AI trained on real legal scenarios. No model, provider, version or hosting location is named, and no commitment to notify customers of a change is published. Speech recognition, the language model and text to speech are usually three separate suppliers in a product like this, and none is identified.

One third party is named in the agreement: customers must accept the Twilio Voice Services Policy and Acceptable Use Policy, which identifies the telephony carrier. Naming who carries the call says nothing about whose model conducts the conversation, so whose system is listening is not stated.

A real figure with its unit is published without a sales conversation, but the agreement contradicts the terms attached to it. LegalClerk's pricing page states 400 dollars per seat per month with no setup fee for unlimited AI calls, alongside a free trial and an enterprise tier at custom pricing, with the features of each published. That is a rate, a unit and an implementation cost. The enterprise tier is withheld entirely.

The published terms also conflict with the agreement. The page says cancel anytime, while section 7.8 bills the service annually in advance with no refunds, section 12.1 renews it automatically unless notice is given thirty days ahead, and section 7.7 sets the free trial at three days against the seven advertised. Usage fees for outbound and toll free calls sit outside the flat rate, and their per minute rates are not published and are available only by email. The agreement also limits unlimited use to 3,000 minutes per user per month.

Practice coverage is documented with real substance, and the boundary is never drawn. Seven practice areas have dedicated material with intake logic for each: personal injury, family law, criminal defense, immigration, real estate, business and corporate, and estate planning. The treatment is operational: accident type detection and liability filtering for injury work, calm structured intake for family matters, and multilingual handling for immigration callers.

Firm size is addressed from solo practitioners to high volume and multi location firms, with multi location team management reserved to the enterprise tier. Jurisdictional pages exist at state and city level. Nothing says where the product stops: no practice area is identified as unsuitable, in house and government legal work are not addressed, no call type is described as one the system should not take, and the languages covered by multilingual intake are not listed.