L
LegalSifter
LegalSifter is an AI contract review company whose software product, ReviewPro, reads a contract inside Microsoft Word or Google Docs, flags risks and missing terms against a playbook using more than two thousand pre-trained machine-learning classifiers the company calls Sifters, and applies generative redlines with one click; lawyer-authored standard playbooks for NDAs, goods and services agreements, master services agreements and SaaS agreements are included and customers can build their own. The company also sells managed services around the software, including playbook customisation, managed redlining, contract data extraction and contract operations as a service, and owns Contract Logix, a contract lifecycle management platform hosted on Microsoft Azure; full-service review by a lawyer is a separate offering from the self-service software. LegalSifter, Inc. is a Delaware corporation headquartered in Pittsburgh, Pennsylvania, operating since 2013 with customers in eighteen countries, and it states plainly that it is not a law firm and does not provide legal advice. ReviewPro pricing is published per contract, at eighty, seventy-two and sixty-five dollars per contract of up to ten pages on the Basic, Team and Enterprise tiers, and organisations can configure whether OpenAI or Google Gemini processes their contracts; the company states that clients specify when their data is deleted and that it will notify clients of a confirmed data breach within twenty-four hours.
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 machine learning is the mechanism the buyer pays for. ReviewPro is priced per contract reviewed, and the review is the product: more than two thousand pre-trained Sifters spot issues and missing terms against a playbook and generative AI applies the redlines; remove the models and there is a playbook document and nothing to run it. The company has sold contract-specific AI since 2014 and the software line is distinct from its human services. Product page, sign-up page and May 2025 launch release read 6 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.
A measured accuracy figure is published, short of a described test set and named failure modes. The free-trial page states 95 per cent plus accuracy in issue spotting, formatting and tone, and the launch release describes contract-specific AI developed and tested over ten years; no test set, method, date or failure-mode statement accompanies the figure on the surfaces read. The primary-authority limbs do not apply to a tool that reads the customer's own contracts against a playbook. Free-trial page and launch release read 6 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.
The modes and the review surface are published, short of the full control structure. The product runs a first-pass review, flags issues with guidance and proposes redlines that a person applies with one click inside Word or Google Docs; the playbook a customer configures sets the rules, positions and fallback language, and the product page states that AI automates routine tasks and highlights key issues for human review. What is not published is a threshold at which any redline is applied without a person or a stated route back after a wrong redline beyond the reviewer's own edit. Product page, sign-up page and Google Workspace listing read 6 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.
Scale and use claims stand in for named deployments on the surfaces read. The vendor states customers and partners in eighteen countries and a decade of contract AI; industry pages exist for healthcare, insurance, manufacturing, technology, higher education, law firms and staffing and were not opened, and no named customer with a figure was located in the material read. A third-party profile refers to university sponsored-research use and is not credited. Product pages and navigation read 6 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.
Confidentiality is addressed at the level of stated practice and the agreement was not read. The security FAQ states that clients may specify when their data is deleted and that content can be separated and managed under differing scenarios such as litigation hold, and the Google Workspace listing states that contract text goes to LegalSifter's servers and may be processed by OpenAI or Google Gemini according to the organisation's configuration. No statement on training use, privilege or work product, or matter-level segregation was located, and the terms and conditions linked in the footer could not be retrieved on 6 September 2026. Security FAQ, Workspace listing and footer read; terms are the rebuttal route.
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 disclaimer is published without the substance the band asks for. The site footer states that LegalSifter, Inc. is not a law firm and does not provide legal advice, and the product is sold to legal and business teams to review contracts without outside counsel; nothing read describes how the product supports a supervising lawyer's competence duties, and the terms and conditions were not retrievable. No jurisdiction limit is named. Footer and product pages read 6 September 2026; the terms are the rebuttal route.
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.
Development claims without a governance framework, testing regime or accountable owner. The launch release states contract-specific AI developed and tested over ten years and the company describes a combined-intelligence approach of AI plus human expertise; no responsible AI framework, ISO 42001 or equivalent, pre-release testing description or statement about uneven output is published on the surfaces read. Launch release and company page read 6 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.
Substantive published policy covering most of the ground, with the security overview not read. Retention and deletion: the security FAQ states that clients may specify when client data is deleted from LegalSifter's systems and that content can be separated for scenarios such as litigation hold. Incident practice: the FAQ commits to notify clients within twenty-four hours of a confirmed breach of network or computing assets giving potential or actual unauthorised access, or of misuse, disclosure or loss of client data. Sub-processors: a sub-processors page exists showing OpenAI, and the Workspace listing names OpenAI and Google Gemini as configurable providers with no local storage. Access control: the Cloud Security overview linked in the footer was not opened and is the rebuttal route. Security FAQ, sub-processors page and Workspace listing read 6 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.
No liability position was located on the surfaces that could be read, and the surface that would carry one could not be reached. The terms and conditions are linked in the site footer and from the Google Workspace listing but did not surface in search and links are not fetchable on this channel; a LegalSifter, Inc. terms of service for the sister product Adams Contracts exists and is not credited for ReviewPro. This records what is locatable on the date and not a finding that no position exists; the terms and conditions are the rebuttal route and would replace this grade on a read. Product pages, security FAQ and footer read 6 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.
Real integrations, documented, with depth described for the two that matter. ReviewPro runs as a Microsoft Word add-in applying tracked redlines in the document, and as a Google Docs add-in whose marketplace listing documents exactly what it does: reads the current document to extract contract text, creates a redlined copy in Drive, transmits content to LegalSifter's servers and stores nothing locally, with the Google permission scopes listed. The product page states connection to existing apps and CLM software without naming them beyond the company's own Contract Logix. Product page and Workspace listing read 6 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.
Cloud delivery is stated and neither tenancy nor region is addressed with substance. Contract content is transmitted to LegalSifter's servers, the company's Contract Logix platform is described as hosted on Microsoft Azure, and the model providers process content according to configuration; no region, tenancy model or residency option is stated on the surfaces read, and the Cloud Security overview was not opened and is the rebuttal route. Workspace listing, G2 vendor text and footer read 6 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.
No certification with scope was located on the surfaces read. A Cloud Security overview is linked in the footer and referenced in the company's Adams Contracts terms, and the security FAQ describes breach notification and deletion practice, but no SOC 2, ISO 27001, auditor or report route appears in the material read; the Cloud Security page was not opened and is the rebuttal route. Security FAQ and footer read 6 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.
Providers are named and the customer chooses between them; models and change notice are not published. The Google Workspace listing states that organisations configure their preferred AI model provider, that ReviewPro supports OpenAI and Google Gemini and routes requests accordingly, and that contract content may be processed by those providers; the sub-processors page shows OpenAI. No specific model, inference location or change-notification commitment is stated. Workspace listing and sub-processors page read 6 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 figure, a unit and a structure are published. The ReviewPro sign-up page lists Basic at eighty dollars per contract with one user and thirty-plus annual reviews, Team at seventy-two dollars per contract with three users and one hundred-plus reviews, and Enterprise at sixty-five dollars per contract with unlimited users, two hundred and forty-plus reviews and single sign-on, all tiers including standard and build-your-own playbooks, with additional reviews addable to any subscription and the unit defined as a contract of up to ten pages at five hundred words a page, longer contracts costing more. Sign-up page read 6 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.
Segment and coverage are described with substance and one limit is stated. Industry pages address healthcare, insurance, industrial and manufacturing, technology and software, higher education, law firms and staffing; standard playbooks cover NDAs, goods and services, master services and SaaS agreements; the company states customers in eighteen countries. The pricing unit's ten-page limit is a stated boundary on contract length; no contract type or jurisdiction is named as unsupported. Navigation, sign-up page and product page read 6 September 2026.
From $80 per contract
- You pay for each contract the software reviews, not per person.
- The smallest plan is $80 a contract for one user and at least 30 reviews a year.
- Bigger plans bring the price down to $72 and then $65 a contract, and the top one lets everyone in the company use it.
- A contract counts as up to 10 pages; longer ones cost more.
- You can buy extra reviews on any plan.
Published per-contract pricing on the ReviewPro sign-up page. Basic: 1 user, 30+ annual reviews, all standard playbooks, build-your-own playbooks, $80 per contract. Team: 3 users, 100+ annual reviews, $72 per contract. Enterprise: unlimited users, 240+ annual reviews, SSO, $65 per contract. All figures based on contract length up to 10 pages at 500 words per page; longer contracts incur additional cost; additional document reviews can be added to any subscription; detailed pricing on request. Entry price is the Basic per-contract rate, the lowest recurring paid rate at the entry tier. Read 6 September 2026.
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?
A published agreement or policy exists and none of it addresses the question either way, or the document that would answer it could not be read and the summary names the retrieval limit. The summary states which shape the silence takes: an improvement right granted that never names training, or no improvement right granted at all.
No located public material addresses whether customer content trains models, and the agreement that would carry a term could not be read. The security FAQ addresses deletion, segregation and breach notice without a training statement, the Google Workspace listing describes transmission to OpenAI or Google Gemini according to configuration without a training statement, and the terms and conditions linked in the footer did not surface in search and are not fetchable from a link on this channel. The terms are the rebuttal route in either direction. Surfaces checked 6 September 2026.
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 the retention window by instruction. The security FAQ states that LegalSifter ensures clients may specify when client data is deleted from its systems and can separate content and manage it under differing scenarios such as litigation hold; the Google Workspace listing states that no data is stored locally by the add-in. No default period and no zero-retention setting is stated, and retention at the model providers is not addressed. Surfaces checked 6 September 2026.
Ethical Walls and Matter Segregation
Does retrieval respect the firm’s ethical walls, or can the model read across them?
Segregation is asserted in public materials with no published detail on how it is enforced.
Segregation is claimed without documentation of a permission model. The security FAQ states that content can be separated and managed under differing scenarios such as litigation hold, and the Enterprise tier adds single sign-on; nothing describes role-based or matter-level walls within a customer or how the review models respect them, and the Cloud Security overview was not opened. Surfaces checked 6 September 2026.
Third Party Request and Subpoena Notice
If someone subpoenas the vendor for a firm’s data, does the firm hear about it first?
No located term or policy addresses third party requests for customer data.
No located public material addresses whether the customer is told when its data is demanded by a third party. The security FAQ commits to breach notification within twenty-four hours, which concerns security incidents rather than legal process, and the terms and conditions, where a compelled-disclosure clause would sit, could not be retrieved on 6 September 2026 and are the rebuttal route. Surfaces checked: security FAQ, sub-processors page, Workspace listing, product pages.
Primary Law Corpus Provenance
Where does the law in this product come from, and does the vendor have the right to use it?
No located public material identifies the corpus behind the product’s answers.
No located public material identifies a legal corpus behind the product's output, and the product is not built on one: the Sifters classify the customer's contract text and the playbooks carry the vendor's or the customer's positions and fallback language, citing no law. The vendor names its chief content officer, Ken Adams, as an author of its contract-language guidance. Product pages checked 6 September 2026.
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.
No located public material addresses whether authority is checked for subsequent history, and the product does not retrieve or cite primary law; its output is flagged issues, guidance and redlines on the customer's contract. Recorded as the honest value for a product without a citator function. Surfaces checked 6 September 2026.
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.
No located public material describes what the Sifters or the redlining model do when they cannot classify a clause with confidence. The product flags terms found and terms missing and applies playbook-driven edits; no abstention path or confidence signal is described. Product pages and Workspace listing checked 6 September 2026.
Fabricated Citation Record
Does a public court record exist involving output from this product?
No court order, opinion or disciplinary record naming this product has been located as of the date shown. This is a statement about the public record, not a finding about the product.
No court order, opinion or disciplinary record naming LegalSifter or ReviewPro was located as of 6 September 2026. The AI Hallucination Cases database maintained by Damien Charlotin was searched on both names together with a general search for court findings; results returned sanctions involving general-purpose chatbots and commentary, none of which names this product. This is a statement about the public record, not a finding about the product; a contract review tool that cites no authority carries a remote exposure on this signal.
Bar Guidance Alignment
Has the vendor engaged in public with the ethics opinions its buyers are bound by?
No located public material engages with bar or ethics guidance.
No located public material names an ethics opinion, bar rule or professional responsibility framework. The footer states that the company is not a law firm and does not give legal advice, and a law-firms industry page exists and was not opened; no guidance from any bar or regulator on lawyers' use of AI is named on the surfaces read. Footer and product pages checked 6 September 2026.
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.
Law firms are a named industry segment and the published position on the bill is a savings claim: the product page states that AI automates routine tasks and reduces reliance on costly external legal services, and the per-contract pricing is itself a cost comparison against outside counsel. Nothing addresses how AI-assisted review is recorded or disclosed on a client's bill. Product page and sign-up page checked 6 September 2026.
Outside Counsel Guideline Readiness
Can a firm get this vendor through a client’s AI clause without a bespoke negotiation?
A current subprocessor or model provider list is published.
The model providers are named on a public surface and a sub-processors page exists. The vendor-authored Google Workspace listing states that contract content may be processed by OpenAI or Google Gemini according to the organisation's configuration, and the sub-processors page on the vendor's site presents OpenAI; the page renders its list as an image with a collapsed FAQ, so whether it carries other processors is not established, and no forwardable disclosure pack or DPA was located. Surfaces checked 6 September 2026.
Court Disclosure Support
If a judge’s standing order requires an AI disclosure, can the product produce one?
No located public material addresses court disclosure or verification certification.
No located public material addresses court disclosure or verification certification of AI-assisted work. The product applies tracked redlines a reviewer accepts, which leaves an edit trail in the document, but nothing states that a record of the model used, sources and human verification can be exported for a court, and the product produces no court-facing work product. Product pages checked 6 September 2026.