N

NexLaw

Each module is model work aimed at a stage of litigation. NeXa researches precedent with source citations, builds arguments and counterarguments, and drafts memos, motions and pleadings. Document Insights summarizes thousands of pages and flags risks, Deposition Insights finds contradictions and key admissions in transcripts, and ChronoVault maps each fact in a case to its supporting evidence on a timeline.

TrialPrep assembles witness outlines and trial materials, and Courtroom Assistant recalls arguments and exhibits in real time. The Professional tier adds a red team that argues opposing counsel's best case, bad fact detection and outcome probability modeling. Matters and files are the only parts that would function without the models, and the terms describe the AI services as machine learning models that produce probabilistic output.

Every research answer links to the controlling authority it relies on, whether case law, a statute or a regulation, so an attorney can open each source before it goes into a filing. NexLaw markets the research as hallucination free with 100% traceability, and a shareholder at Marquis Aurbach Coffing credits it with 100% citation accuracy on medical record summaries.

No test set, error rate or method stands behind the accuracy claims, and the retrieval method and research sources are not described. Output is described in the terms as probabilistic, and assessing its accuracy for each use falls to the user.

Users and their end users are accountable under the terms for every decision made, advice given and action taken on the strength of the AI output, which is described as probabilistic and to be assessed for accuracy. ChronoVault ties each fact to its evidence, and research answers link to their authority, which gives a reviewer something to verify against.

No review step, approval point, threshold or route back to an attorney is described for drafts, red team analysis or outcome modeling, and nothing says what happens when the courtroom assistant surfaces a wrong exhibit or argument.

Nine case studies name the attorney and firm, each published with the client's consent. Randi S. Sirlin of Sirlin Law Firm in Phoenix cut drafting of complex family law pleadings from 8 to 10 hours to 4 to 5 using NeXa Deep Research, Build an Argument and the pleadings tool. The Scott Practice reports cutting discovery auditing time by more than 90% in estate litigation. Christian T. Balducci, a shareholder at Marquis Aurbach Coffing, says medical record summaries went from three days to 15 minutes.

Other studies cover Southern Arizona Legal Aid, Williams Oby Law Group, The Hingerty Law Firm, Steele Law Corporation, EMC Law and Resnick & Louis. The before and after figures come from the attorneys' own accounts of their hours, the studies carry no dates, and each warns that individual results vary.

Privilege gets its own section in the trust center. Client and matter data is not used to train public models, and each firm works in a logically isolated environment with access controls against leakage across matters and clients. Firms set their own retention rules, and role based access limits who inside the firm can see or act on a matter. NexLaw says it is built around ABA Model Rule 1.6, offers an ethics opinion from outside counsel on request, and includes a data processing agreement in every deployment.

The published terms are broader than these commitments. They license user content to maintain the service and to improve NexLaw's products and create aggregated, deidentified information, and they let staff reach content to comply with legal requirements. Zero data retention is described for enterprise users only, and the data processing agreement is not published.

The terms make users and their end users accountable for all decisions made and advice given on the basis of the AI services, and describe output as probabilistic. They do not say that NexLaw gives no legal advice.

The product is sold to litigators, from solo practitioners to trial teams, yet the about page also invites individuals seeking legal clarity. No disclosure for people who are not lawyers is published, and no jurisdiction limit is stated beyond the focus on litigation in the United States.

Neither an AI policy nor a governance framework, accountable owner or testing program is published. The trust center's section on AI transparency covers the site's llms.txt file and how crawlers may use NexLaw's web content, not how its models are tested. Nothing addresses uneven output across practice areas, jurisdictions or parties, including for the outcome modeling and red team features.

Legal data is kept for seven years or as the law requires under the privacy policy, and firms can set retention rules so matters and AI outputs are kept or deleted in line with their own policies. Data is encrypted with TLS in transit and AES 256 at rest in databases and object storage, keys are rotated under internal policy, and requests are authenticated and authorized before processing.

Change management, access management and incident response procedures are described as formally defined and tested, with continuous monitoring and regular vulnerability assessments. Audit logs record logins, matter actions including each AI analysis run, and permission changes, and administrators can export them. No subprocessor list or breach notice period is published.

The terms of service, last updated 30 July 2025, exclude liability for indirect, incidental, special, consequential and punitive damages, including lost profits and data, and state no cap on direct damages. The service is provided as is with all defects, and merchantability, fitness, non infringement and error free operation are disclaimed.

Users carry responsibility for every decision and action based on the AI output. No indemnity runs from NexLaw to the customer, and no warranty on output or insurance is offered. Annual contracts paid in installments stay payable in full if cancelled early, under the refund and cancellation policy.

NexLaw runs as a standalone web platform. Files are uploaded through its own secure uploader, smaller firms are offered its built in document management, and audit logs export for internal review. No connection to a case management system, a document management system, Microsoft Word or a court filing system is named, and bespoke configuration at onboarding is described without listing what it connects to.

All data is stored and processed on servers in the United States, with no cross border transfers, on what NexLaw calls a private cloud built to store and protect client data. Each firm has a logically isolated environment, and the engineering team configures the deployment to the firm's data environment before it goes live.

Every plan runs on that same deployment in the United States. The cloud provider is not named, and no on premises option or choice of region is described, although the demo form takes requests from the United Kingdom, Canada, Australia and New Zealand.

SOC 2 Type II certification, independently audited across the security, availability, confidentiality and privacy criteria, is claimed, with the full report available under NDA on request. The trust center describes the same program as aligned with SOC 2 Type II and the infrastructure as aligned with SOC 2, without naming the auditor, the review period or the scope. GDPR and HIPAA compliance are listed as badges.

The AI services are described as machine learning models that generate probabilistic predictions from patterns in data, and the trust center separates NexLaw's environment from public AI services. No model, provider, host or region for the models is named, nothing says whether the models are NexLaw's own or licensed, and no notice of a change in models is promised.

Prices are published in USD as annual licenses per user, with no minimum seat count. Essential costs $2,999 a year for solo practitioners and boutique litigation firms, covering research, IRAC structured drafting, deposition analysis and document review. Professional costs $3,999 a year and adds ChronoVault chronologies and timelines, red team stress testing, bad fact detection, outcome modeling, volume seat licensing, dedicated training and a custom retention and DPA arrangement.

The license includes onboarding, updates, standard support and bespoke configuration, and NexLaw states there is no implementation fee. A five day trial is offered, and qualifying firms can run a pilot scoped to one practice group, ending in an ROI report. Annual contracts paid by installments remain payable in full after cancellation.

Litigation in the United States is the focus, for solo practitioners, boutiques, trial teams and high volume litigators, with enterprise plans adding SSO. Practice areas named include personal injury, medical malpractice, criminal defense, family law and appeals, and estate litigation, and the case studies add legal aid and multistate practice. Solution pages cover discovery, litigation strategy and contested contract and breach analysis.

Outside litigation, coverage is not described, and no practice area or jurisdiction is stated as unsupported, although demo requests are taken from the United Kingdom, Canada, Australia and New Zealand.