V
Visalaw AI
Visalaw AI is a research and drafting platform built for immigration lawyers by Visalaw Ventures, Inc. of Memphis, Tennessee, a company formed by people at the immigration firm Siskind Susser to take the firm's internal tools to the wider immigration bar. It answers immigration research questions with citations from a curated library of agency materials from DHS, USCIS, DOJ, DOS, DOL and HHS, AILA publications and a case law collection of more than 10 million cases, and drafts petitions, briefs and support letters for a published list of business, family and humanitarian case types, from O-1A and EB-1A to asylum, U-visas and BIA appeals.
It also analyses uploaded documents, translates and summarises them, assembles exhibit lists, and organises work by project with team permissions and audit trails. It is designed for attorneys who are AILA members, with paralegals working under a supervising attorney's account. Visalaw publishes its full contract set, including a platform agreement, data processing addendum and security addendum that bar training on customer data by Visalaw or its model providers. Published plans start at $220 a month, with Pro at $480 and custom Enterprise pricing.
Capability grades
All 15 axes, graded from public sources on the date shown. 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 drive the research and drafting the vendor leads with, on a platform with real non-AI functions, which is the B band. Natural-language research over the immigration library, petition and brief drafting, legal analysis, document chat, translation and summarisation are model work. Around them sit projects organised by matter, team permissions, audit trails, e-signature, a certified-translation option and a large curated library of agency materials, AILA content and case law that has value in its own right. Verified 18 September 2026.
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.
Grounding is real and documented against a named corpus with citations on every output, short of any measurement, which is the B band. The platform page describes the research library, agency materials from DHS, USCIS, DOJ, DOS, DOL and HHS, AILA secondary sources and more than 10 million cases from over 2,000 courts searchable by jurisdiction, and the FAQ says every output includes transparent citations so attorneys can cross-check.
No accuracy figure, evaluation or failure mode is published, nothing says what the system does when the library does not support an answer, and the FAQ's claim that the product ensures accuracy every time is not backed by any measurement. The Platform Agreement states that output is AI-generated and may contain errors. Verified 18 September 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.
A written position that the attorney stays in control, with real review surfaces and a supervision rule for staff, short of a published control structure, which is the B band. The Platform Agreement calls the service a research and drafting tool whose output is not legal advice and makes the customer responsible for evaluating output for accuracy; the about page says technology should support professional judgment rather than replace it.
Paralegals and assistants may use the product only under a supervising attorney's account, with audit trails, and drafts are edited in the platform before use. No mode, threshold or categorical limit on what an output may be used for is published. Verified 18 September 2026.
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 customers with some figures, short of any method, which is the B band. The customer stories page names Charles Kuck of Kuck Baxter, Cyrus Mehta of Cyrus D. Mehta & Partners, Chase Hertel of SimpleCitizen, and solo practitioners John Manley and Ken Robinson, who says the platform saves him four to six hours a week and let him cut his hourly rate. The platform page claims up to 90 per cent reductions in drafting time and an ROI of 10 hours per case and $60,000 a year, stated as based on an assumption of ten drafts a month. None of the figures is measured across a described sample. Verified 18 September 2026.
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.
Every element of the confidentiality picture is contractual, and the vendor addresses privilege and work product directly, which is the A band. The Platform Agreement commits that Visalaw will not train any AI model on customer data or content and that its model providers will not either, that subprocessors other than cloud storage will not retain or log content for human review, and that customer content is the customer's confidential information for the term and ten years after.
The Security Addendum documents per-organisation isolation enforced at the API and database layers, including in the shared vector index, and contractually bars model providers from training, from keeping content beyond a single API call and from sharing it. The Data Processing Addendum limits de-identification to operational metadata and says matter-level content never leaves the customer's tenant. The vendor's March 2026 post on United States v. Heppner explains why the court denied privilege and work product protection to a client's consumer AI output, sets out the Kovel route for AI used at counsel's direction, and advises attorneys to use enterprise tools with contractual no-training and confidentiality terms and to document attorney direction, pointing to Visalaw's own commitments. Flagged for sampling as the pull's first Privilege A. Verified 18 September 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.
A clear position on what the product is, who may use it and how staff use is supervised, short of jurisdiction limits, which is the B band. The Platform Agreement states that the service is a research and drafting tool and that output is not legal advice; the platform FAQ says the product is designed for immigration attorneys and professionals who are AILA members, and that paralegals and assistants may use it only under a supervising attorney's account with audit trails.
No jurisdictional limit is stated, and the same FAQ asserts that the product ensures ethical compliance and meets professional responsibility standards, a claim a tool cannot make on a lawyer's behalf. Verified 18 September 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.
Values and AI-specific data controls are published without a governance framework for model behaviour, which is the C band. The about page states principles, including that every answer must be grounded in the law, and the Security Addendum has a dedicated AI Model Security section on provider restrictions, tenant isolation and data flow documentation; the DPA commits to renegotiate if new AI regulation requires it. Those are security and data commitments rather than governance of model behaviour: nothing names who owns model quality, what is tested before release, or what has been found about uneven output across case types, languages or nationalities. Verified 18 September 2026.
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 stewardship set is published in contract detail, with the subprocessor list itself unreadable, which holds this at the top of the B band. The Security Addendum commits to AES-256 encryption at rest and TLS 1.2 or higher in transit, least-privilege access with multi-factor authentication, annual penetration tests, stated vulnerability remediation windows, background checks, and notice of a security incident within 24 hours with an independent forensic assessor at Visalaw's expense.
The DPA commits to return or delete customer data within 30 days of termination with written certification, and to 60 days' notice of new or changed subprocessors with objection and termination rights. The subprocessor list is held in the Trust Portal, hosted on Drata behind an access request, which refuses automated reading, so the top band is not asserted on it. Verified 18 September 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.
A detailed published liability position with named insurance, short of anything that stands behind output, which is the B band. The Platform Agreement gives a vendor indemnity against third-party claims that the service infringes intellectual property or arises from Visalaw's violation of law, caps general liability at twelve months' fees with carve-outs for confidentiality breaches and indemnities, sets a separate cap at twice annual fees, and commits Visalaw to maintain technology errors and omissions and cyber liability insurance, with certificates on request. Insurance amounts are not stated, and output accuracy is placed on the customer. Verified 18 September 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.
Named integrations across immigration practice systems, short of documented depth, which is the B band. The platform page lists integration partners including the eImmigration case management system, Clio, Docketwise, Candle AI for email drafting, MotaWord for translation, and Google Drive, OneDrive and Dropbox, and says research and drafts fit case management and email workflows; API developer access is included on the Pro plan.
What each integration exchanges and in which direction is not documented on the surfaces read. Verified 18 September 2026.
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.
Tenancy, hosting region, residency options and where processing happens are all stated, which is the A band. The Security Addendum states that customer data and content are stored and processed by Visalaw and its vendors in U.S. data centres, hosted by default in the AWS and MongoDB US East 1 regions with replication only within the United States, that customers may request a specific region, and that the service runs on shared infrastructure with logical separation enforced by organisation at the application and database layers; model providers receive content over encrypted API calls and may not keep it beyond the call.
Encryption keys are provider-managed, with no customer-managed keys. The DPA describes the same arrangement as a dedicated tenant environment, a difference in wording recorded here. Verified 18 September 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.
A named attestation with a contractual route to the report, short of evidence a reader can see, which is the B band. The Security Addendum states that the security programme is audited against SOC 2 Type II, that ISO 27001 is targeted for the fourth quarter of 2026, and that customers receive recent SOC 2 reports, penetration test summaries and questionnaire responses on request at no cost; the security page adds ISO/IEC 42001 as in progress, a dedicated CISO, SAML single sign-on, audit logs and IP allow-listing.
The auditor, report period and scope are not stated on any page read. The Trust Portal is hosted on Drata behind an access request, and the host refuses automated reading, so its contents could not be read. Verified 18 September 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.
The AI architecture and provider obligations are described with change notification, without the providers named on any surface read, which is the B band. The Security Addendum describes content passing to contracted model providers over encrypted APIs, bars them from training, retention beyond a single call and onward sharing, and says Visalaw documents which subprocessors receive data at each stage; the Platform Agreement and DPA give 60 days' notice before a new or changed subprocessor, with objection rights, and the security page says these obligations extend to external model providers.
The providers themselves are listed only in the Trust Portal, which is hosted on Drata behind an access request and refuses automated reading; a 2024 announcement co-issued with AILA named GPT-4, which is dated and not relied on. Verified 18 September 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.
A buyer can price the product and see what each plan includes before a sales conversation, which is the A band. The pricing published on the platform page sets Core at $220 a month for research, the curated library, document analysis, translation, projects, skills and case management integration, and Pro at $480 a month adding up to ten petition and brief drafts a month, custom skills, exhibit review, collaboration and API access, with Enterprise custom for unlimited drafting.
Onboarding differs by plan, with customised onboarding on Pro and Enterprise and in-person training on Enterprise, and qualifying firms on annual plans receive two free support-staff seats per attorney licence. Enterprise rates are withheld. A pricing row is written. Verified 18 September 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.
Segments, practice areas and limits are documented precisely, which is the A band. The product is for immigration attorneys who are AILA members, from solo practitioners to mid-to-large firms, and the platform page lists the case types supported for drafting: business (O-1A, O-1B, EB-1A to EB-1C, NIW, H-1B, TN, L-1, E-1 to E-3, R-1, I-360), family (I-601, I-601A, I-485, N-400, N-600), humanitarian (asylum, T and U visas, VAWA, cancellation of removal) and multi-category work such as RFEs, motions and BIA appeals.
The research library is scoped to named agencies and courts, with separate U.S., Canadian and global libraries. Anything outside immigration is outside the product by design. Verified 18 September 2026.
7 public documents
The public pages on file for Visalaw AI, with the recorded signals each one supports and the date it was last read. Open any of them and check the reading against the record.
-
visalaw.ai/platform-agreement2 signals
Client Data in Training, Prompt and Output Retention
Read Sep 18, 2026
-
visalaw.ai/platform2 signals
Primary Law Corpus Provenance, Court Disclosure Support
Read Sep 18, 2026
-
Third Party Request and Subpoena Notice
Read Sep 18, 2026
-
visalaw.ai/security-addendum1 signal
Ethical Walls and Matter Segregation
Read Sep 18, 2026
-
Fabricated Citation Record
Read Sep 18, 2026
-
Bar Guidance Alignment
Read Sep 18, 2026
-
visalaw.ai/customer-stories1 signal
Billing and Fee Posture
Read Sep 18, 2026
$220 / monthUSD, as published, never converted
- The Core plan costs $220 a month for immigration research, document analysis and translation.
- The Pro plan costs $480 a month and adds drafting of up to 10 petitions and briefs a month.
- Big firms can get an Enterprise plan with unlimited drafting at a custom price.
- Firms on yearly plans can get two free seats for support staff per attorney.
- You need to be an AILA member to use it.
Core: $220 per month; research and AI chat, curated immigration dataset, AILA content access (membership required), document analysis, summarisation and translation, projects, skills library, case management integration, standard support. Pro: $480 per month; adds up to 10 petition and brief drafts per month, custom skills, exhibit list review, legal analysis, in-platform editing, collaboration and e-sign, Drafts pilot, API access, customised onboarding.
Enterprise: custom; unlimited drafting, prompt engineering support, tailored and in-person training. Qualifying firms on annual plans receive two free support-staff seats per attorney licence.
Note: Figures from the pricing section of the vendor's platform page, read 18 September 2026. The page does not state whether plans are priced per attorney licence; the site banner refers to attorney licences on annual plans.
Legal Signals
What each signal meansA signal records what public sources say on the date shown. It is not a grade and it is not a recommendation. Where a signal reads Not addressed, it means the index did not locate the material in public sources on that date, which is a statement about disclosure rather than about the product.
Client Data in Training
Can material a lawyer puts into this product be used to train a model?
The published terms prohibit training on customer content. Not a policy page, the agreement.
The published Platform Agreement prohibits training on customer data and content, and extends the bar to the vendor's model providers. Section 4.3 commits that neither Visalaw nor its third-party model providers will train any AI model on customer data or content, and that subprocessors other than cloud storage will not retain or log it for human review; the Data Processing Addendum repeats the commitment and says it is enforced through the subprocessor agreements.
Prompt and Output Retention
How long does the product keep what a lawyer typed, and can that be set to zero?
The customer controls the retention window, by product configuration or by contractual instruction, but zero retention is not stated as available.
The customer controls return or deletion of its data, and model providers keep nothing beyond the call. The Platform Agreement and DPA commit to return or delete customer data and content at the customer's election within 30 days of termination, with written certification, and the platform describes data as user-managed. The Security Addendum bars model providers from keeping content beyond a single API call. No shorter in-term retention setting for prompts and outputs is described.
Ethical Walls and Matter Segregation
Does retrieval respect the firm’s ethical walls, or can the model read across them?
The product maintains its own permission model, documented, requiring the firm to keep it aligned.
The product's own separation model is documented. The Security Addendum says every record is scoped to an organisation identifier enforced through access tokens at the API layer, and that embeddings in the shared vector index are filtered by organisation and draft; the platform offers projects organised by matter with private team access, role-based permissions and audit trails. Walls between matters inside one firm rely on project permissions, which are described but not documented in detail.
Third Party Request and Subpoena Notice
If someone subpoenas the vendor for a firm’s data, does the firm hear about it first?
Terms commit to notice where lawfully permitted. No transparency report located.
The Data Processing Addendum commits to notify the customer of legally binding law enforcement requests for its data, to the extent legally permitted, and to use commercially reasonable efforts to obtain a waiver where notice is prohibited. It also commits to advance notice where law requires processing outside the customer's instructions. No transparency report was located.
Primary Law Corpus Provenance
Where does the law in this product come from, and does the vendor have the right to use it?
Sources are identified without stating the licence or rights basis.
Sources are named by publisher and scale, with curation described, without a licence basis for each. The platform page names agency materials from DHS, USCIS, DOJ, DOS, DOL and HHS, AILA publications and practice pointers, the vendor's own content, and a case law library of more than 10 million cases from over 2,000 courts, curated daily; access to AILALink and AILA.org content requires AILA membership. The rights basis for the case law collection and AILA material is not stated.
Good Law Verification
Does the product tell you when the authority it just cited has been overruled?
No located public material addresses whether authority is checked for subsequent history.
Searched the platform page and FAQ, the legal documents and the customer stories on 18 September 2026. The product cites case law from a library of more than 10 million cases, but nothing states whether cited cases are checked for subsequent history or flagged when overruled.
Refusal and Uncertainty Behaviour
What does the product do when the answer is not in the corpus?
No located public material addresses what the product does when it cannot ground an answer.
Searched the same surfaces on 18 September 2026. No abstention or uncertainty behaviour is described. The FAQ says every output carries citations for cross-checking, which supports verification but does not describe what the system does when the library lacks support.
Fabricated Citation Record
Does a public court record exist addressing fabricated or hallucinated legal citations in output from this product?
No court order, opinion or disciplinary record addressing fabricated or hallucinated legal citations produced by this product has been located as of the date shown. This is a statement about the public record on that one subject, not a finding about the product, and this signal is not a litigation history.
Searched the AI Hallucination Cases database maintained by Damien Charlotin and trade press reporting on 18 September 2026 for court or tribunal records addressing fabricated or hallucinated citations in output from Visalaw AI or Gen. None located. This signal does not record litigation history of any other kind.
Bar Guidance Alignment
Has the vendor engaged in public with the ethics opinions its buyers are bound by?
Public materials refer to professional responsibility in general terms without naming guidance.
The vendor names a specific court decision on AI and applies it to practice. Its March 2026 post on United States v. Heppner explains Judge Rakoff's ruling that a client's consumer AI output was neither privileged nor work product, discusses the Kovel doctrine for AI used at counsel's direction, and advises immigration attorneys to counsel clients in writing, use enterprise tools, document attorney direction and adopt a firm AI policy.
The platform FAQ adds that paralegal use runs under a supervising attorney. No bar ethics opinion or EOIR guidance is named on the surfaces read.
Billing and Fee Posture
Does the vendor address what happens to the bill when the work takes an hour instead of six?
Public materials claim time savings without addressing billing or disclosure, and the product sits inside a fee relationship between a lawyer and a client where those savings would change the bill.
Time and cost savings are claimed for work inside a lawyer and client fee relationship, without billing guidance. The platform page claims 10 hours saved per case and $60,000 a year on an assumption of ten drafts a month, and a customer story reports a solo attorney who cut his hourly rate by $300 because of the time saved. Nothing addresses how firms should bill or disclose AI-assisted work.
Outside Counsel Guideline Readiness
Can a firm get this vendor through a client’s AI clause without a bespoke negotiation?
This signal has not been recorded for this vendor yet. It is not a finding either way.
Court Disclosure Support
If a judge’s standing order requires an AI disclosure, can the product produce one?
Some elements of the record are available, short of a document level export.
Some elements of a record exist, short of a document-level export. Every output carries citations to its sources, and the platform keeps audit trails of team activity. Nothing records which model produced a draft or who verified it, and no disclosure export is described for filings with USCIS, EOIR or courts.