InPractice vs Superinsight: how they compare in 2026
InPractice and Superinsight both read medical records and turn them into chronologies and summaries for claims lawyers and medical examiners. InPractice is stronger on checking. It sets each generated summary beside the original record, links chronology entries to their source, and says human review is required before final use. Its privacy policy lets personal information be processed anywhere in the world. Superinsight is stronger on its privacy notice and practice depth. Its HIPAA notice commits that health information is not shared with AI or language model companies, and it builds reports for disability, veterans, injury and mass tort claims. It states that no human reviews the records, and its terms cap its liability at $250. InPractice charges by the page, and Superinsight by the credit, with about two credits per report.
Which one fits
Choose InPractice if
- You want to check every entry against the record. InPractice shows each generated summary and chronology entry beside the original record, links entries to their source document, and lets you edit before export to Word, Excel or PDF.
- You want to pay only for pages processed. InPractice sells page credits with no subscription or seat fee, from $100 for 1,000 pages, credits never expire, and a trial covers 500 pages without a card.
- You bill record review to the case. InPractice produces a per case cost statement of pages used and total spend, exportable to PDF for internal tracking or case billing.
Choose Superinsight if
- You handle disability or veterans claims. Superinsight builds residual functional capacity assessments and evidence reports for Social Security Disability and VA disability, with separate treatment for workers' compensation, mass tort including Camp Lejeune, and malpractice.
- You want health information kept from model vendors in writing. Superinsight's HIPAA notice, effective January 2026, commits that protected health information is not shared with AI or large language model companies except as required by law or with authorization.
- You want tiers you can compare. Superinsight publishes six options from $80 for one credit to $5,000 a month for 200, states that a report averages about two credits, and gives the first case free.
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.
Every function is model output: chronology construction, narrative summarization, OCR search, duplicate detection and the Case Chat feature that answers questions against the record set. The product is billed per page processed, with no seat fee and no subscription, so a customer pays for inference directly. The vendor describes itself as specializing exclusively in medical record review rather than offering a broad legal AI platform. Without the models there is a file store with a search box.
The models are both the product and the billing unit. Every output is generated by a model, from chronologies and narrative summaries to research queries and residual functional capacity assessments. Charging is by credit consumed per report, roughly two credits a report, so a customer pays for inference and nothing else. The pipeline is described as fully automated, with no person reviewing the records. Remove the models and there is an upload form.
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.
Chronology entries carry an outline view that jumps directly to the underlying source document, and a side by side review places each generated summary and chronology entry next to the original medical record for verification and editing. Export can optionally include the original records, so the chain from output back to source survives outside the platform. The vendor's published limitations state that the product requires human oversight for final verification and that AI generated content needs review. No accuracy figure, test set or evaluation is published; the one published number, three seconds per page, measures throughput.
Accuracy is asserted through outcome claims, and no grounding mechanism is described on any surface read. The claims are specific in form and unsupported: 3x more evidence found than manual review, 70 percent time savings on the home page against up to 90 percent in the FAQ on the same page, and 80 percent cost reduction. None has a test set, method, sample or date, and the two time savings figures contradict each other. Nothing describes how an output traces back to the page it came from, and no citation, source link or verification view appears on the home, pricing or legal pages.
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.
Generated summaries and chronology entries sit next to the original records in the side by side view, and every entry is individually editable before export, so the user controls what leaves the system. The vendor's published limitations state that human oversight is required for final verification and strategic decision making. Duplicate detection operates unattended and is described as hiding exact duplicates only. No threshold is published at which the system defers or flags low confidence, nothing describes what happens when an error survives review, and there are no terms of service in which a supervision obligation would sit.
The process runs in three steps, upload, analysis by AI, then a report, and the FAQ says it requires no human review of sensitive medical data, framed as a privacy advantage. Nothing sits between the record set and the report except the model. Clause 3.1 of the terms says the user assumes all risks of using content, including reliance on its accuracy. The marketing site and the terms describe no review surface: no verification view, no confidence signal, no threshold at which the system defers, and no statement that a lawyer must check output before use. The documentation site has separate pages on building reports, managing reports and research insight.
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.
Five testimonials carry both a person and an organization: Wayne Winders of Williams and Swee, Dylan Barket of Barket Lawyers, Simone Sacks of Orthopedic Expert Services, George Gluck MD of the Hand Center of Nevada, and Jaclyn Meeks of AmarilloMD, spanning the legal and medical examiner segments. Winders reports that case preparation time was cut by half and that demand letters improved because detail buried in hundreds of pages was no longer missed. The vendor separately claims use by hundreds of firms and organizations nationwide. The single figure is self reported, with no matter volume, period or method, nothing is dated, and no case study is published on the site.
Five attributed testimonials are published, each with a photograph, full name, role, organization and practice area, among them Katie Reed of McMahan Law Firm, Scott Haider of Schneider Law Firm and Paul Bunn of Veterans Outreach Ministries. Bunn's carries a figure tied to his organization: the product saved about 300,000 dollars a year and let him replace three registered nurses and two nurse practitioners while keeping review quality. The company separately claims more than 500 attorneys and hundreds of firms. Nothing is dated, no case study is published, and the corporate percentages are not attached to any named customer.
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 privacy policy states in its second paragraph that it does not apply to protected health information processed on behalf of customers, which is governed by the business associate agreements InPractice has with its customers, and those agreements are not published. The input to the product is medical records, so the published policy covers account and marketing data. The security page states that customer information is never used to train AI models or shared with third parties, and describes 256-bit AES at rest, AES encryption in transit and multi factor authentication by SMS or TOTP. Nothing addresses privilege or work product, nothing describes separation between customers or matters, and no retention or deletion position is published. No model provider is identified, so what any underlying provider may retain is unstated.
The published HIPAA notice answers directly what an outside model provider may see. The HIPAA Notice of Privacy Practices, effective 12 January 2026 and public rather than held behind a contract, states that Superinsight does not share personal information, including protected health information, with vendors performing services for it. It names artificial intelligence and large language model companies and cloud and database integration services, except as required by law or with explicit authorization. Clause 3.2 confines the content license to returning results to the user. Nothing addresses privilege or work product or describes separation between customers or matters, and no retention period exists.
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.
The site publishes no terms of service, customer agreement or acceptable use policy: the navigation carries Product, Industries, Security, Pricing and Company, and the footer carries a privacy policy alone. Nothing states that output is not legal advice or not medical advice, and no attorney client or clinician patient boundary is addressed. The same product is sold to attorneys, to qualified and independent medical examiners who write expert reports, and to insurers adjusting claims, and its output is a medical chronology used to argue causation. The published limitations state that human oversight is required for final verification and strategic decision making, which concerns output quality rather than the advice boundary.
No position on the advice line is published, and the audience makes that gap a practical one. The terms of use, the privacy policy and the HIPAA notice were read in full, and none states that output is not legal advice, not medical advice or not a substitute for professional judgment. The product is sold to lawyers, to independent and qualified medical examiners producing expert evaluations, and to representatives who are not lawyers. Two of the five published testimonials come from consulting and outreach practices working veterans disability claims, not law firms.
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 responsible AI statement, governance framework, named owner accountable for model behavior, description of evaluation before release or testing result is published. Nothing addresses bias: a chronology tool determines which diagnoses, gaps and preexisting conditions surface, and uneven performance across record types, providers or handwriting quality would change what a reader concludes about causation. The Vanta partnership named on the security page automates compliance monitoring for HIPAA, a security function.
An AI standard is invoked with no framework published behind it. The home page states the platform is aligned with ISO 42001, alongside HIPAA compliance. ISO 42001 is an AI management system standard that an accredited body can certify against, and a claim of alignment is the company's own assessment, not an audited certification. Nothing behind the claim is published: no management system scope, no owner accountable for model behavior, no evaluation regime and no test results. Nothing at all addresses bias, on a product that decides which clinical findings support a disability or causation argument.
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 it does not apply to customer protected health information, the material the product processes. Published access controls are 256-bit AES at rest, AES encryption in transit and multi factor authentication supporting both SMS and TOTP. Infrastructure suppliers are named as AWS, Google Cloud Platform and Microsoft Azure, and Google Analytics and Google Ads are named in the privacy policy. No retention period is published for uploaded records, summaries or chat content, no deletion commitment or route exists, and no incident or breach notification practice is published. The privacy policy provides that personal information may be transferred, processed and stored anywhere in the world.
Some safeguards are published, and those governing the records after processing are not. The HIPAA notice commits to breach notification in line with HIPAA requirements and gives rights to inspect, amend and restrict records. Access control is described as a one time passcode sent to the user's phone, plus encryption and firewalls. The records have no published retention period or deletion route. The security language is dated: the privacy policy describes secure socket layer technology and was last updated on 21 August 2023, more than two years before the HIPAA notice it sits alongside.
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 terms of service, master subscription agreement, customer agreement or order form template is published; the navigation has no legal section and the footer's only legal link is the privacy policy. No indemnity, liability cap, warranty of any kind, service level or insurance position is published. The pricing page invites buyers to purchase page credits self-serve with a credit card. The business associate agreements referenced in the privacy policy govern data protection rather than liability for output, and are unpublished.
Liability is limited by standard clauses that disclaim the exposure the product creates. Clause 7 limits liability for any damages arising from or related to the terms to a maximum of two hundred and fifty US dollars, and states that more than one claim does not enlarge the limit. Clause 6 provides the app as is, with all warranties disclaimed, accuracy included. Clause 4 runs the indemnity one way, from user to company. Under clause 5.3 the user releases and forever discharges the company from every claim relating to the app, including personal injuries and death, and California residents waive Civil Code section 1542. Two published items point the other way. Clause 2.4 commits the company to keep the app in good working order and promptly correct material defects that materially impact the delivery or accuracy of reports. The home page offers a money back guarantee described as carrying no risk. Both are service commitments, not allocations of loss. No insurance position is published.
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.
No case management system, document management system, medical record retrieval service, intake platform or electronic signature product is named, and there is no integrations page, API reference or developer documentation in the navigation or footer. Chronologies and summaries export to Word, Excel or PDF with a choice of fields, filters and layouts and the original records optionally included, which moves a finished document out of the platform; records arrive by upload. The vendor states that the product is not an EHR or EMR system.
The documentation site has an Integration page under Manage Case. No case management, document management, intake, medical record retrieval or electronic signature product is named on the home page, the pricing page or the legal documents, and the marketing site has no integrations page or API reference. Records arrive by upload.
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.
Three cloud providers are named on the security page, AWS, Google Cloud Platform and Microsoft Azure, and no region, country or data center is given. Nothing describes whether the platform is single or multi tenant, and no dedicated, private or isolated option is offered at any tier. The privacy policy provides that all personal information processed may be transferred, processed and stored anywhere in the world, including but not limited to the United States or other countries with different privacy laws, for a platform whose stated market is the United States only and whose content is protected health information.
Neither tenancy nor region is stated. No hosting provider, cloud platform, country, region or data center is named on any surface read, and nothing says whether the platform is single or multitenant. No dedicated, private or isolated option appears at any of the six pricing tiers. The HIPAA notice and the privacy policy are both silent on where protected health information is stored or processed, and the privacy policy has no international transfer section. The one nearby statement is the export control clause in the terms, which refers to US export laws. Searches covered the site and all three legal documents.
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.
HIPAA compliance is claimed, and HIPAA has no certification scheme; the security page describes a partnership with Vanta, a compliance automation vendor whose product monitors controls, so the Vanta HIPAA badge evidences a monitoring subscription rather than an audit. No SOC 2 of any type, ISO certification, auditor, report date, scope statement or penetration testing is published, there is no trust center, and no report is offered at any access tier, including on request.
HIPAA compliance is claimed repeatedly; HIPAA has no certification scheme, so the claim rests on the vendor's word. ISO 42001 is claimed as alignment, the vendor's own wording, not certification. The site claims no SOC 2 of any type, names no auditor or certifying body, and gives no report date, scope statement or penetration testing. A Trust Cloud portal at superinsight.trustshare.com is publicly linked from the home page and the FAQ. Two further badges are not security credentials: NVIDIA Inception is a startup program and the CIO Applications listing is a magazine award.
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.
No model is named, no model provider is identified, no subprocessor list exists, no location is given for inference, and no commitment to notify customers of a change is published. AWS, Google Cloud Platform and Microsoft Azure are named on the security page as the cloud platforms the product runs on. The security page says customer information is never shared with third parties, and nothing says whether inference runs on a proprietary model or a third party one.
No model, version, provider or hosting location is named, and the published statements say only who does not receive the data. Both privacy documents rule out passing data to AI or language model companies. No commitment to notify customers if the models or their providers change is published. What reads the records, beyond the company's own AI, is not identified.
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.
Pricing is published in schedules that disagree. The volume table gives 1,000 pages at 0.10 dollars a page for 100 dollars, 5,000 at 0.09 for 450, 20,000 at 0.07 for 1,400, 100,000 at 0.06 for 6,000 and 200,000 at 0.05 for 10,000. The four purchase cards below it give Starter at 100 dollars for 1,000 pages at 0.10, Pro at 400 dollars for 5,000 at 0.08, Growth at 1,500 dollars for 25,000 at 0.06 and Scale at 5,000 dollars for 100,000 at 0.05. Three of the four overlapping tiers disagree on both total and rate, and the cards offer a 25,000 page tier the table does not contain. The separate AI information page gives a third version whose 200,000 page row reads 0.10 a page against a 10,000 dollar total, which does not compute. Only the 1,000 page entry tier at 100 dollars is stated the same way everywhere. Every tier includes unlimited users, chronologies, AI summaries, export, OCR and support, with no implementation charge, and the 500 page trial over seven days needs no credit card.
Pricing is complete enough to choose a tier and start without speaking to anyone. Six options are published with figures, the unit is defined as a credit with a report averaging about two, and a table sets out what a credit costs on every plan. What changes between tiers is itemized, credit expiry rules are stated, and new customers get their first case without charge or a card. A short questionnaire recommends a tier.
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.
Three buyer segments each have their own page: legal, medical examiners and insurance. Practice areas are named as personal injury, workers' compensation, medical malpractice and mass tort, and the medical examiner audience is identified as qualified and independent medical examiners scaling their practices. The vendor states what it does not do: it specializes exclusively in medical record review rather than general legal AI, it is not an EHR or EMR system, and it serves the United States market only. Firm size is not addressed, in house legal departments and government use are not mentioned, and nothing states a record volume limit or the document types the system cannot process.
Practice coverage is detailed, and the boundary is left open. Eight practice areas each have a page describing an analytical approach rather than a label, such as residual functional capacity for Social Security Disability, service connection for veteran disability and standard of care deviations for medical malpractice. Mass tort names Camp Lejeune specifically. Firm size is addressed through the pricing tiers, sized by monthly case volume from one or two cases up to fifty to a hundred or more. State pages are published for all fifty states. The limit is missing: use by corporate legal teams or government is not addressed, and a Custom Solutions page offers to build for any practice area not listed instead of saying where the product stops.
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 security page states without qualification that InPractice never uses customer information to train AI models or shares it with third parties, and that the data belongs to the customer, with no exception for deidentified or aggregated data. The statement sits on a marketing surface: the site publishes no terms of service, customer agreement or data processing agreement, and the privacy policy states that it does not apply to customer protected health information, pointing instead to business associate agreements that are not published.
Training is not addressed either way, and two published provisions limit use without naming it. Clause 3.2 of the terms of use lets the company use customer content solely to provide information back to the user, with all other rights kept by the user; that is a purpose limit, not a prohibition. The HIPAA notice's bar on sharing with AI and language model companies concerns who receives the records, not what is done with them.
Neither document says whether Superinsight trains its own models on customer content, and no marketing page addresses it. This reflects a full reading of all three legal documents on 4 September 2026.
Prompt and Output Retention
How long does the product keep what a lawyer typed, and can that be set to zero?
No retention period is published for uploaded records, summaries, chronologies or Case Chat content: the privacy policy has no retention section, the security page carries no retention statement, and there are no terms of service. No deletion commitment, deletion route or end of engagement position is published, and nothing states what happens to a case file once page credits are exhausted or an account goes dormant. Page credits never expire.
The agreement addresses retention without stating a period. Clause 3.1 says the company is not obligated to back up user content and that user content may be deleted at any time without prior notice, which puts the backup burden on the customer. Clause 8 adds that terminating an account may involve deleting user content from live databases, leaving any other copies unaddressed. No retention window applies during use, no deletion request route is described, and no certificate of deletion is offered.
Ethical Walls and Matter Segregation
Does retrieval respect the firm’s ethical walls, or can the model read across them?
Nothing states whether the platform is single or multi tenant or how one customer's records are separated from another's. Within an account, credits can be shared across team members and used across cases, users are unlimited at every tier, and the privacy policy records that a customer may access information associated with a user's use of the Services, including the contents of files associated with that account. No role model, permission structure or matter level access control is described, and multi factor authentication is the only access control published.
No role or permission model is described on the marketing site or in the three legal documents, and they do not say how one customer's records are kept apart from another's or address walls at matter level. The documentation index lists pages for sharing a case, managing groups and managing organizations.
Third Party Request and Subpoena Notice
If someone subpoenas the vendor for a firm’s data, does the firm hear about it first?
The privacy policy reserves the right to disclose personal information as required by law and where InPractice believes disclosure is necessary or appropriate to protect rights, property or safety, enforce policies or contracts, assist an investigation, or comply with a judicial proceeding, court order or legal process served on it. No commitment to notify the customer of such a request appears, and no discretion over notice is reserved either.
The policy states at the outset that it does not apply to customer protected health information, so this provision governs account and marketing data, and nothing published addresses compelled disclosure of the medical records. No transparency report is published.
The HIPAA notice provides for disclosure of protected health information when federal, state or local law requires it, including to the Secretary of Health and Human Services for compliance investigations, and separately for public health and safety. The privacy policy allows sharing personal information with outside parties on a good faith belief that disclosure is reasonably necessary to meet legal process or an enforceable governmental request.
Neither document commits to advance notice or reserves discretion over it. After the fact, the HIPAA notice gives a right to request an accounting of disclosures made for purposes other than service provision. No transparency report is published.
Primary Law Corpus Provenance
Where does the law in this product come from, and does the vendor have the right to use it?
InPractice reads the medical records the customer uploads to a case and produces chronologies and summaries from them, rather than retrieving external content. No external database, publisher, clinical coding set or reference source is named, and no license basis is stated.
The output is built from the medical records a customer uploads, not from external content, and no source collection is identified. Two outputs relate to published frameworks, residual functional capacity assessments and evidence matched to disability listings, but no source, edition, publisher or license basis for any framework is named. No clinical coding set or reference database is identified either. This was checked against the site pages and legal documents on 4 September 2026.
Good Law Verification
Does the product tell you when the authority it just cited has been overruled?
The product does not retrieve or present primary law: it summarizes and sequences the medical records supplied for a case, so no citator applies, and none is claimed.
The product does not retrieve or present case law, so checking authority for later history does not arise. It analyzes the medical records supplied for a claim and produces chronologies, summaries and evidence assessments, and nothing on any located surface addresses the question. Searches on 4 September 2026 covered the site's main and practice area pages and the legal documents.
Refusal and Uncertainty Behavior
What does the product do when the answer is not in the corpus?
The vendor's published limitations state that AI generated content needs review and that human oversight is required for final verification. No abstention path, no-answer state or confidence or grounding score surfaced to the reviewer is described, and nothing states how the system handles illegible handwriting, poor scans or gaps in a record set.
Nothing located describes what the product does when it cannot ground an output. The published workflow has no checkpoint between upload and report. No abstention path, no state for declining to answer and no confidence or grounding score shown to the user is described. Nothing addresses the ordinary failure conditions for this kind of product, such as illegible handwriting, poor scans or providers missing from a record set. As of 4 September 2026, neither the site pages nor the legal documents described any of this.
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 InPractice or InPractice, Inc.
The AI Hallucination Cases database kept by Damien Charlotin holds no entry naming Superinsight as the tool used. Superinsight builds chronologies, summaries and assessments from medical records for claims representation, including veterans disability and Social Security disability cases.
Bar Guidance Alignment
Has the vendor engaged in public with the ethics opinions its buyers are bound by?
No bar association, regulator, rule of professional conduct or ethics opinion is named on the home page, security page, pricing page, privacy policy or AI information page, and there is no terms of service or acceptable use policy. A blog post cites American Bar Association survey data on the rate of AI adoption among legal professionals, as a source of statistics.
No bar association, regulator, rule of professional conduct or ethics opinion is named in the legal documents or on the main site pages. Two of Superinsight's customer segments, veterans disability and Social Security disability representation, are governed by their own federal accreditation and fee approval regimes. The product generates evidence arguments and capacity assessments for those claims, and nothing published addresses either regime.
Billing and Fee Posture
Does the vendor address what happens to the bill when the work takes an hour instead of six?
Case Cost Statements are described as clear reporting of page usage and total spend per case, exportable to PDF to support internal tracking or case billing. The product is AI processing charged per page, so every figure on that statement records AI assisted work on that matter: the pages counted are the pages the model read. Nothing addresses how that cost should be disclosed to a client or described on an invoice, no guidance on fee treatment is published, and nothing addresses what happens to an hourly entry when review time falls from days to minutes.
Savings are claimed, and billing and disclosure are not addressed. Nothing connects the cost and time figures on the home page, or a named customer's report of about 300,000 dollars saved a year, to a client bill or fee petition. No record of AI work for each matter is described. Disability and veterans representation are contingency and fee approval practices, where the cost of preparing evidence is itself a regulated question, and nothing published engages that.
Outside Counsel Guideline Readiness
Can a firm get this vendor through a client’s AI clause without a bespoke negotiation?
No subprocessor list is published, no model provider is identified, and no data processing agreement, consent pack or notification template exists at any access tier. AWS, Google Cloud Platform and Microsoft Azure, named on the security page, are infrastructure. Business associate agreements are referenced in the privacy policy as governing customer protected health information, are not published, and no route to obtain one before contracting is described.
The HIPAA notice is the one forwardable item, and the lists a client would ask for are not published. Drafted for individuals, the notice answers the model provider question in the negative, and any disclosure to a business associate requires written agreements and appropriate safeguards. No subprocessor register exists, no model or provider is identified, and no consent or notification pack is published. The Trust Cloud portal is linked as the route to further security detail.
Court Disclosure Support
If a judge’s standing order requires an AI disclosure, can the product produce one?
Chronology entries link to the underlying source document, and the side by side view pairs each generated entry with the original record. Export can optionally include the original records with the chronology, and Case Cost Statements record pages processed and spend for each matter. Nothing states that the model behind a given entry is recorded or disclosed, no exportable log of which entries a person edited or approved is described, and no disclosure guidance or template for a court or an opposing party is published.
The outputs are the work product itself, chronologies, summaries and assessments built from templates, and the legal documents and the home and pricing pages describe no export, audit trail or log of how they were produced. No model is recorded against an output, there is no human verification step to record, and no disclosure guidance or template for a tribunal is published. In Superinsight's main practice areas, a representative may need to account to the Social Security Administration or the Department of Veterans Affairs for how evidence arguments were prepared.
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.
- UPL and Professional Responsibility Posture
- Deployment Model and Data Residency
- Ethical Walls and Matter Segregation
- Primary Law Corpus Provenance
- Good Law Verification
- Refusal and Uncertainty Behavior
- Bar Guidance Alignment
In summary
InPractice
InPractice, from InPractice, Inc. of Goleta, California, founded in 2023, reads medical records and turns them into chronologies and narrative summaries. It serves personal injury, workers' compensation, medical malpractice and mass tort practices, medical examiners and insurers in the United States. It processes records at about three seconds a page and offers Case Chat and export to Word, Excel or PDF. According to the AI Legal Index, InPractice pairs each summary with its source record and states that human review is required before final use. It states it does not train on customer information and prices from $100 per 1,000 pages.
Superinsight
Superinsight, based in Menlo Park, California, reads medical records and produces chronologies, summaries and case reports for claims that turn on medical evidence. It gives separate treatment to Social Security Disability, personal injury on both sides, veterans disability, workers' compensation, mass tort, malpractice and medical examiners. According to the AI Legal Index, Superinsight's HIPAA notice commits that health information is not shared with AI companies, and it states that no human reviews the records. It publishes six price options, from $80 for a single credit to $5,000 a month for 200 credits.
Questions buyers ask
InPractice vs Superinsight: which is better for medical record review?
InPractice is stronger on verification, pairing each summary with its source record. Superinsight is stronger on its written commitment about health information and on practice specific reports for disability and veterans claims. Disability practitioners have more to use in Superinsight. From the AI Legal Index, based on each vendor's own published materials as of September 27, 2026. No vendor pays for placement.
Do InPractice and Superinsight share records with AI companies?
Superinsight's HIPAA notice commits that protected health information is not shared with AI or large language model companies except as required by law or with explicit authorization. InPractice states that customer information is never used to train AI models or shared with third parties. From the AI Legal Index, based on each vendor's own published materials as of September 27, 2026. No vendor pays for placement.
How much do InPractice and Superinsight cost?
InPractice charges by the page with no subscription, from $100 for 1,000 pages, though its larger tiers appear at different prices in different places on its site. Superinsight charges by credit, from $250 a month for 4 credits or $80 for a single credit, with a report averaging about two credits and the first case free. From the AI Legal Index, based on each vendor's own published materials as of September 27, 2026. No vendor pays for placement.
What happens if the output is wrong?
Superinsight's terms disclaim accuracy, cap its liability at $250 and run the indemnity from the user to the company, while committing to correct material defects affecting report accuracy. Both place review of the output on the user. From the AI Legal Index, based on each vendor's own published materials as of September 27, 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.
Three points to weigh. InPractice publishes no terms of service, its privacy policy excludes the medical records it processes, and its pricing page shows schedules that do not agree beyond the 1,000 page tier. Superinsight's terms cap liability at $250 and disclaim accuracy, its ISO 42001 claim is alignment rather than certification, and its home page states two different time saving figures. Neither vendor reviewed this page.