A
ArcPrime

ArcPrime is an AI patent platform and IP management system for in house patent teams, built to replace a legacy IPMS rather than sit beside one. Jon Liu, a former head of patent analytics at Meta, founded the company in 2025, and it announced $5.1 million in combined seed and preseed funding led by RTP Global and Rackhouse Venture Capital. The system ties invention disclosures, prosecution records, prior art, portfolio decisions and spend together.

An inventor's email can become a structured disclosure, and patent counsel can review similarity, filing risk and fit with a business or product taxonomy. It holds patent and trademark records and brings in prosecution events and documents from patent office sources.

AI workflows find prior art, draft first responses to office actions for outside counsel to refine, map claim elements to a product or reference, and recommend candidates for continuation or pruning. Staff review the sources, steer the analysis and decide which results to use. Reporting covers portfolio coverage, spend and outside counsel. A REST API and an OAuth secured Model Context Protocol server let authorized tools read or update workspace records within assigned permissions.

ArcPrime runs on Microsoft Azure OpenAI by default and offers self hosted open source models for enterprise deployments. Subscriptions come with unlimited users and either metered or unlimited AI usage, and implementation typically takes 2 to 6 weeks.

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.

AA on AI CentralityThe artificial intelligence is the product. Remove the models and there is nothing left to sell.

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.

AI runs through each stage of the patent work alongside the portfolio and docket. It turns inventor submissions into structured disclosures, finds prior art, ranks filing candidates on criteria the legal team sets, and prepares strategies and draft responses to office actions. After grant it builds claim charts, runs deep research on companies practicing a claim, and suggests where to continue or prune coverage, with pruning sessions showing cost, a patent rating and savings to date.

Each task draws on the records and instructions in the workspace. Patent records and deadlines form a conventional system of record, and AI analysis and drafting sit at the center of the patent workflows.

Source: Vendor Published
BB on Citation Accuracy and Hallucination DisclosureGrounding is real and documented, with linked primary sources and a described retrieval method, short of published accuracy figures an outsider can test.

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.

Prior art results list references a reviewer can open, and the assistant shows sources under its answers. Claim charts map each claim element to text in a source document, and counsel can review the references and reasoning behind a proposed office action response before using it. Outputs are meant for professional review. There is no representative citation error rate and no independent verification method for every cited authority.

Patent office status updates and links to prosecution records do not show whether a legal authority is still current or has been followed by a court.

Source: Vendor Published
BB on Autonomy and Oversight ModelA written commitment that the models work alongside a supervising lawyer, with real review surfaces, short of the full control structure: commonly the threshold at which the system stops or what happens after it is wrong.

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 patent team sets the weights for filing recommendations, reviews the factors behind an assessment and steers office action drafting, which produces a first draft for outside counsel to refine. The system can prepare analysis or send reminders automatically when a workflow condition is met, and authorized connected assistants can write to the portfolio within their permissions. Users are told to review AI outputs before use, which keeps counsel in charge of decisions and filings.

No mandatory human approval applies to every connector write, and no rule says when an uncertain response must stop for review.

Source: Vendor Published
BB on Operational and Outcome EvidenceReal deployment evidence with substance, short of full attribution or measurement: a named customer without figures, or figures without the named customer.

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.

The company counts in house patent teams at Fortune 500 companies, Fortune 10 leaders among them, as users. Its case studies leave customers unnamed. A semiconductor leader managing tens of thousands of patents cut renewal fees while keeping strategic coverage, and a global software provider handling hundreds of disclosures a year reports faster fact gathering and patent coverage mapped to revenue streams. Testimonials come from roles such as the head of IP at a large enterprise SaaS provider and at a large financial services company. None names a customer or measures accuracy or speed against a baseline.

Source: Vendor Published
BB on Privilege and Confidentiality PostureSubstantive published commitments on confidentiality and training use, short of the full picture: commonly silence on segregation between users or matters, or on what the underlying model provider may retain.

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.

Each tenant's data is separated at the application and storage layers. Patent documents, disclosures and generated outputs stay in the workspace under encryption and role permissions, customer data does not train models, and support access is restricted and logged. Forms can include questions visible only to internal legal staff or to authorized outside counsel. The DPA is available to current and prospective customers under NDA. No privilege waiver undertaking is offered, and matter level separation is not described for every connected AI system.

Source: Vendor Published
BB on UPL and Professional Responsibility PostureA real position is published on advice versus tooling, short of full treatment: commonly a disclaimer without the supervision and competence dimension, or silence on jurisdiction limits.

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 tool prepares work for patent professionals, including prior art analyses, claim charts and draft office action responses. The help center tells users to check outputs before finalizing work and treats them as needing professional judgment. Counsel review reasons and source documents rather than taking a ranking as the answer. Workflow and form permissions limit what an inventor or outside attorney sees. Nothing maps the workflow to individual bar opinions or blocks a request from someone not authorized to give legal advice.

Source: Vendor Published
BB on AI Governance and Bias DisclosureA published governance framework with real substance, short of testing results or a named owner.

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.

Patent teams can adjust the weights behind filing recommendations and add factors tied to their own strategy. The tool gives reasons for individual factors and shows sources under its answers. Role controls and the usage log let an administrator see who can act and which case each AI job ran against. Testing for bias across technical fields or patent offices is not described, and model changes have no stated approval process. The controls center on how a customer constrains and inspects the tool inside its own patent process.

Source: Vendor Published
BB on AI Safety and Data StewardshipSubstantive published policy covering most of the ground, short of the full set: commonly no named subprocessor list or no stated incident practice.

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.

Customer data never trains ArcPrime's AI models, and AI results stay inside the tenant under standard workspace controls. Data is deleted at contract termination under the company's retention policy, with a copy available for export beforehand. Encryption uses AES-256 at rest and TLS 1.3 in transit, audit logging runs across every customer environment, and support access is restricted and logged. No retention period is set for AI prompts during the subscription, no backup purge schedule is given, and no rules cover an outside AI client reaching the same workspace through the connector.

Source: Vendor Published
CC on AI Liability and RecourseLiability is addressed only through a standard limitation clause that disclaims the exposure the product creates.

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.

The site terms, effective April 16, 2025 and issued by ArcIP, Inc., govern the website, and the Services they mention are never defined. They disclaim all warranties, exclude indirect and consequential damages without a dollar cap, and make the user indemnify the company, including for third party intellectual property claims. Disputes go to binding AAA arbitration with a class action waiver under California law. The DPA and related materials sit in the trust center behind an NDA, and no subscription agreement in view sets rights for an error in a prior art report or an office action draft, damages under a software order, or recourse for an incorrect recommendation.

Source: Vendor Published
AA on Practice Systems Integration DepthDocumented, verifiable integrations into the systems legal work already lives in, with the depth described: what syncs, in which direction, and what a firm must configure.

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.

USPTO legal events sync into portfolio records, and documents from foreign offices arrive through Global Dossier. Migration runs from Anaqua, IPfolio, CPA/Memotech, Equinox, MaxVal, PatSnap and Questel, and the platform connects to electronic billing, HR and inventor records, and identity providers through SAML 2.0 and OAuth. A REST API and an OAuth secured Model Context Protocol server expose portfolio data to authorized applications, and with write permission those tools can update patents, add notes or create projects. Migration fills the system at the start, while ongoing record updates and connector permissions handle the daily work.

Source: Vendor Published
BB on Deployment Model and Data ResidencyDeployment model is stated clearly with partial residency detail, or residency is offered without the processing location being addressed, or the tenancy model is stated on its own with no residency detail published.

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.

ArcPrime runs as a hosted application with tenant separation at the application and storage layers, on servers on the West Coast of the United States, with Supabase listed as the cloud provider. The privacy policy says all information is processed and stored in the United States. Azure OpenAI is the default AI platform, and enterprise deployments can use self hosted open source models, which covers model processing rather than putting the portfolio database on customer infrastructure.

Portfolio hosting has no selectable regions, and data sent to outside connected models and applications carries no residency commitment.

Source: Vendor Published
AA on Security Certifications and Trust CenterCurrent independent attestation with named scope, reachable without a sales call: a trust center carrying reports, dates and the standards actually covered.

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.

SOC 2 Type 2 and ISO 27001:2022 certifications are held, with an annual independent audit, and the trust center also lists GDPR. The SOC 2 Type 2 report, the ISO audit report and the ISMS certificate are available on request, along with a security questionnaire, subprocessor list, DPA and policy library, to current and prospective customers under NDA. Controls include single sign on, role permissions, workspace isolation, AES-256 encryption at rest and TLS 1.3 in transit, and support access is limited to authorized staff and logged. Buyers can request the documents to confirm dates and scope for the services they plan to use.

Source: Vendor Published
AA on Model Supply Chain DisclosureThe models underneath are named, their providers identified, where they run is stated, and the vendor commits to notifying customers when any of that changes.

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.

Microsoft Azure OpenAI is the default AI platform, and enterprise deployments can run self hosted open source models instead. Some tasks vary the model, such as deep research, which is priced per target and model. The workspace also connects to outside AI clients such as Claude, ChatGPT, Microsoft Copilot and Gemini through an OAuth secured Model Context Protocol server, kept separate from the default inference path.

The trust center's subprocessor list names GitHub, Google Workspace, Supabase, Vanta and Cursor, with no model provider among them. No full list ties model versions to each task, and routing rules for each configuration are not described.

Source: Vendor Published
BB on Commercial TransparencyReal pricing is published for part of the range, with enterprise tiers withheld, or the unit and structure are stated without the figure.

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.

Subscriptions are quoted and come with unlimited users, which the company contrasts with per seat pricing for legacy IPMS tools. AI usage is either unlimited or metered in units allocated by plan. At standard rates, finding prior art and proofreading drafts are included, an office action response recommendation costs 250 units with free reruns on the same action, and a continuation recommendation is included on a subscription or costs 10 units otherwise.

Deep research is priced per target and model. A dialog shows the cost and balance before each run, and contracts can set different rates. No dollar price applies to the subscription, the units or migration.

Source: Vendor Published
AA on Firm and Practice CoverageWho the product serves is documented precisely: firm segments, in house and government use, and the practice areas actually supported, with the limits stated.

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.

The product serves corporate patent departments managing inventions, active prosecution, patents and trademarks across business units. It connects to USPTO records with automated synchronization and brings in documents for selected European and Asian filings through Global Dossier. Teams use it for prior art, prosecution drafting, examiner analytics, maintenance and continuation choices, competitive and licensing intelligence, docketing, and spend or coverage reports by business area.

Outside counsel can work inside it with authorized access. Case studies span a semiconductor company, an enterprise software provider, a precision health company, a robotics and AI company and a consumer electronics company. The platform stays focused on intellectual property and its budget workflows rather than every legal practice.

Source: Vendor Published
Sources on file

7 public documents

The public pages on file for ArcPrime, 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.

Pricing

No published figure

  • Subscriptions are quoted and include unlimited users.
  • AI usage is unlimited in some workspaces and metered in units in others.
  • At standard rates, finding prior art and proofreading drafts are included, an office action response recommendation costs 250 units, and a continuation recommendation is included on a subscription or costs 10 units otherwise.
  • Deep research is priced per target and model.
  • Implementation typically takes 2 to 6 weeks.

Quoted workspace subscription with unlimited users. Metered plans start with a unit allocation and can buy more units, and contracts can include features or set different rates.

Implementation: The company manages migration, validation, setup and training, typically in 2 to 6 weeks, with no separate fee schedule.

Note: Units carry no dollar value, and a confirmation dialog shows each run's cost and the balance before it starts. The subscription itself is quoted.

Legal Signals

What each signal means

A 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.

Confidentiality and Privilege

Client Data in Training

Can material a lawyer puts into this product be used to train a model?

Never, in policy only

A public policy or trust page states no training on customer content, with no matching term located in the published agreement.

The security page and homepage say customer data is never used to train ArcPrime's AI models. The website terms carry no training clause, and the DPA is available only under NDA.

Source: Vendor PublishedCustomer data is never used to train ArcPrime’s AI modelsAs of Nov 19, 2026Evidence

Prompt and Output Retention

How long does the product keep what a lawyer typed, and can that be set to zero?

Disclosed without a period

Retention is acknowledged in public materials with no stated period.

AI outputs stay in the workspace, and data is deleted at contract termination under the retention policy after an optional export. No fixed period covers prompts or backups.

Source: Vendor Publishedupon contract termination in accordance with ArcPrime’s data retention policyAs of Nov 19, 2026Evidence

Ethical Walls and Matter Segregation

Does retrieval respect the firm’s ethical walls, or can the model read across them?

Own model, documented

The product maintains its own permission model, documented, requiring the firm to keep it aligned.

Customer data is segregated per tenant at the application and storage layers, with no cross customer model or index, and role permissions extend to connected tools.

Source: Vendor PublishedYour data is segregated at the application and storage layer per tenant.As of Nov 19, 2026Evidence

Third Party Request and Subpoena Notice

If someone subpoenas the vendor for a firm’s data, does the firm hear about it first?

Disclosure addressed, notice absent

Published terms or policy address disclosure to authorities or in response to legal process, and no commitment or reservation regarding customer notice is located anywhere. The vendor has told the customer that data can leave and has said nothing about whether the customer hears of it.

The privacy policy allows disclosure to government or law enforcement officials in response to legal requests, including court orders and subpoenas. Affected customers are not promised notice.

Source: Vendor Publishedto respond to legal requests (including court orders and subpoenas)As of Nov 19, 2026Evidence
Accuracy and Authority

Primary Law Corpus Provenance

Where does the law in this product come from, and does the vendor have the right to use it?

Sources named, basis unstated

Sources are identified without stating the license or rights basis.

USPTO data with automated synchronization and office documents from USPTO Global Dossier are the named sources. License terms for the full prior art collection are not identified.

Source: Vendor PublishedAs of Nov 19, 2026Evidence

Good Law Verification

Does the product tell you when the authority it just cited has been overruled?

Not addressed

No located public material addresses whether authority is checked for subsequent history.

The system retrieves prosecution events and prior art. No case citator or judicial treatment check is part of the product.

Source: Operator VerifiedAs of Nov 19, 2026Evidence

Refusal and Uncertainty Behavior

What does the product do when the answer is not in the corpus?

Confidence signal only

The product exposes a confidence or grounding score without an explicit abstention path.

Filing recommendations come with risk and factor assessments that a team can reweight. No rule holds back an uncertain recommendation.

Source: Vendor PublishedAs of Nov 19, 2026Evidence

Fabricated Citation Record

Does a public court record exist addressing fabricated or hallucinated legal citations in output from this product?

None located

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.

The court AI hallucination tracker holds no case tying fabricated citations to ArcPrime.

Source: Operator VerifiedAs of Nov 19, 2026Evidence
Professional Responsibility

Bar Guidance Alignment

Has the vendor engaged in public with the ethics opinions its buyers are bound by?

Not addressed

No located public material engages with bar or ethics guidance.

Patent professionals are told to verify outputs and apply professional judgment before use. The workflow is not tied to specific bar guidance.

Source: Operator VerifiedAs of Nov 19, 2026Evidence

Billing and Fee Posture

Does the vendor address what happens to the bill when the work takes an hour instead of six?

Audit record only

A usable record of AI assisted work exists with no published fee guidance.

The usage log records case, task type, units spent and date, with a link to each result. Outside counsel spend runs as its own workflow, and AI fee disclosure to clients has no guidance.

Source: Vendor PublishedAs of Nov 19, 2026Evidence

Outside Counsel Guideline Readiness

Can a firm get this vendor through a client’s AI clause without a bespoke negotiation?

On request only

The material exists behind a sales conversation or an executed agreement.

Subprocessors are GitHub, Google Workspace, Supabase, Vanta and Cursor, and the DPA, the audit reports and a security questionnaire go to customers under NDA.

Source: Vendor PublishedAs of Nov 19, 2026Evidence

Court Disclosure Support

If a judge’s standing order requires an AI disclosure, can the product produce one?

Partial record

Some elements of the record are available, short of a document level export.

AI tasks are logged with the case, task type, units, date and a link to the result. No court disclosure export carries each prompt and model setting.

Source: Vendor PublishedAs of Nov 19, 2026Evidence
Contact

Correct a record, or ask how something was graded

Every grade and every signal on this index is drawn from public sources and dated. If a record is wrong, out of date, or missing an artifact the index did not locate, send the source and it will be reviewed and the record redated. Vendors are welcome to submit documentation. Nothing on this index is for sale, including a listing, a placement, or a grade.

AI Legal Index

The AI Legal Index is an independent index that tracks changes to AI vendors in legal. It holds 535 vendors across 9 categories, each graded on the same 15 capability axes and recorded against 12 legal signals, from public sources. No vendor pays for inclusion, placement, or rating.

Index Status
Last index update
November 19, 2026
The AI Legal Index is an editorial reference. It is not a regulatory body, not a law firm, and nothing published here is legal advice or a recommendation to retain or avoid a vendor. Records are verified against published sources, bar guidance and public court records. Where a record reads not addressed, the material was not located in public sources on the date shown. See the Methodology page for evaluation standards and limitations.
© 2026 AI Legal Index
3801 N Capital of Texas Hwy, Ste E240 · Austin, TX 78746