R
Red Points
Red Points is an AI brand protection platform that detects and removes counterfeit listings, fake websites, impersonating profiles, fraudulent ads and pirated content on a customer's behalf. Its models scan more than 5,000 platforms across marketplaces, social networks, search engines, app stores, domains and mainland China platforms, using image, logo, shape, packaging and text recognition trained on each protected brand asset, and score the seller behind a listing rather than the listing alone across a database of more than 17 million seller identities.
Enforcement runs through direct API integrations and more than 300 removal agreements with platforms, and every detection and decision is logged with what triggered it and packaged as evidence formatted to the infringement type, platform and jurisdiction. The platform is sold in tiers from Starter to Enterprise across five programmes covering counterfeits, gray market and unauthorised distribution, brand impersonation, domain abuse and piracy, charged as a flat fee for an agreed scope with unlimited detections and takedowns inside it.
Expert-led services can be added for online investigations, test purchases, IP registration, domain dispute resolution and a zero-cost litigation programme for recovering funds from infringing sellers. Buyers are the brand protection, intellectual property, legal, e-commerce and security functions of consumer and industrial brands, and Red Points reports more than 1,300 of them. The company is independent, was incorporated in Barcelona as Red Points Solutions, S.L., and has offices in New York, Barcelona, Beijing and Salt Lake City.
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 vendor's own framing is stronger than the band it earns: the Our AI page, modified 23 July 2026, states that AI is no longer the tool, it is the operator, and describes a five-layer Digital Risk Operating System of autonomous agents, governed deterministic logic, forward-deployed engineers, a brand knowledge layer of encoded judgement and a data bedrock of 2.7 billion data points analysed monthly. Detection, prioritisation, seller scoring and evasion-resistant image matching are all model work and none of it survives removing the models.
What does survive is the reason this is B rather than A: the enforcement estate is not AI. More than 300 removal agreements, direct API connections with named platforms, priority reporting and escalation paths, the IP document repository and the case workflow would all still function, and an analyst-driven enforcement desk built on exactly that substrate is what the legacy vendors in this market sell. The models are the engine of detection layered on a workflow and channel platform that would still operate without them. Verified 12 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.
R15 governs: this product cites no legal authority, so the primary-authority grounding and citator limbs do not apply to the product class and are neither credited nor penalised. What bites is grounding and accuracy disclosure. Grounding is real and documented in the product's own idiom: the vendor publishes that every decision is logged with what was detected, why, what triggered it and what happened next, that Vision AI shows the exact logos, patterns or products that triggered a match, and that evidence is generated automatically and formatted to each infringement type, platform and jurisdiction, so a reviewer opens the triggering artifact rather than a summary of it.
What keeps this off A is that no measurement is published anywhere located: no precision, no recall, no false-positive rate, no test set, and no third-party validation, while the marketing does make accuracy claims, including that the AI eliminates false positives and that models are custom-trained per asset and overseen daily. The published figures are economic rather than accuracy figures: 10.3 times lower average cost per enforcement against analyst-based models, and 12.6 full-time-equivalents of average capacity per brand protected. Verified 12 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 published control structure is more explicit than most records in this corpus and still short of the A band on the two limbs the B band names. What runs alone is stated plainly: autonomous agents handle prioritisation, tagging, enrichment, rule suggestion, documentation and reporting and, in the vendor's words, run the workflow rather than assisting it, escalating to a person only when it is required. What constrains it is a stated two-tier split, with a governed tier described as deterministic, fully auditable logic reserved for validation, enforcement and legal-ready evidence, the decisions the vendor says cannot go wrong.
Review surfaces are real: live and custom dashboards, per-decision logs, generated evidence packages and a document repository, with a dedicated customer success manager and IP-operations specialists named in the plan tiers, and human-in-the-loop oversight sold as a feature rather than a slogan. Two limbs are missing. The threshold at which an agent escalates instead of acting is never published beyond the phrase when it is required.
And nothing located addresses what happens after the system is wrong: a validated detection that removes a legitimate seller's listing has no published remediation, appeal or withdrawal path on any vendor surface read. Verified 12 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 and figures, attributed to each other, and undated with no method, which lands on B. Published on the vendor's own pages with a name, a company and a title: James Larson, VP and Deputy General Counsel of IP at Purple, against 4,700+ monthly enforcements; Satoru Ishiguro, Head of IP Division at Topcon, against $22M in infringements removed; Marta Benedetto, General Manager at FILA Italy, against 638,653 listings removed; Stephan Jacob, Founder and CGO at Cotopaxi, against 4,700+ fake websites removed and 130+ hours reclaimed in a quarter; Jennifer Carman Parker, Manager of IP and Global Brand Protection at KEEN; plus Kaspersky at $1.6M in infringements enforced and Burton at 4,600+ fake websites removed, each with its own case study page.
That is fuller attribution than the B band's stated shape. What holds it off A is that not one figure carries a date and no method is stated for any of them, and the monetary figures are the ones that most need one: nothing published explains how $22M or $1.6M of infringements removed or enforced was valued. Analyst placement (Gartner and G2 ratings) is recorded rather than credited, because ratings are not deployment evidence. Verified 12 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.
This record contains the most direct statement on privilege located anywhere in this corpus, and it is a disclaimer. The Legal Notice and Terms of Use states that Red Points is not a law firm, that its employees will not act as the customer's attorney, that no attorney-client relationship is created, and that while communications with Red Points are covered by its Data Protection Policy they are not protected by the attorney-client privilege or work product doctrine.
An IP counsel sending infringement evidence and enforcement instructions into this platform has been told in writing that the exchange is not privileged, which is a material fact for the buyer and is the reason this row is worth reading rather than skipping. Everything else on the axis is generic or absent. The customer agreement is not published at all: the same Legal Notice states that any provision of services is formalised individually through channels outside the site, so the terms that would govern training, retention, deletion and segregation cannot be read before signing.
The Data Protection Policy, last updated July 2019, is a website and marketing policy whose collection channels are contact forms, the newsletter, the blog, demo requests, corporate email, social profiles and the careers page. The Security Policy is a statement of corporate security principles and objectives that speaks to protecting the confidentiality of the company's own communications. Confidentiality is therefore asserted in general terms and the operative commitment lives in a sales conversation, which is the C band on both of its limbs. Verified 12 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.
An unusually full published position on the advice line, and one that goes further than most vendors in this index are willing to. The Legal Notice and Terms of Use states that Red Points is not a law firm, does not provide legal advice, does not intend to engage in the practice of law, and provides only self-help services at the customer's specific direction; that it gives no advice, explanation, opinion or recommendation to a consumer about legal rights, remedies, defences, options, selection of forms or strategies; that no attorney-client relationship arises; and that the customer is representing itself in any legal matter it undertakes through the services.
It also states that the intellectual property information it publishes is not legal advice and is not guaranteed to be correct, complete or current, because the law changes quickly, differs between jurisdictions and is interpreted differently by different courts and authorities. Two things keep this at B. Nothing addresses the competence and supervision duties of the lawyer who is in fact the buyer on many of these accounts, which the B band names as the common shortfall.
And R50 applies to the instrument: this is a website legal notice, not the service agreement, though its disclaimer section speaks to Red Points' services and employees by its own words rather than only to the site, which is why it grades here at all. Verified 12 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.
What is published is an architecture claim rather than a governance position. The Our AI page names a Governed AI tier described as deterministic, fully auditable and traceable across every validation and enforcement workflow, grounding every decision with evidence, logic and accountability, and states that every decision is logged with what was detected, why and what triggered it. That is real product auditability and it is recorded here.
None of the A or B limbs is met. No accountable owner inside the vendor is named, no pre-release testing regime is described, no evaluation results are published, and there is no AI policy, model card, review board or published framework of any kind. Most notably for this product class, nothing is published about uneven output: the system makes consequential decisions about third parties rather than about the buyer, since a validated detection removes a named seller's listing and a Seller Risk Score labels a trading entity as a repeat infringer, and no error rate, no appeal statistic and no analysis across seller types, categories or jurisdictions is disclosed. Verified 12 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.
A generic privacy policy covers the company without addressing what the platform does with the material it holds, which is the C band. The Data Protection Policy, last updated July 2019, enumerates its collection channels as the website and contact forms, the newsletter, the blog, cookies and advertising partners, demo requests, corporate email, social profiles and the careers page, and states its processing purposes as answering queries, sending news and information of interest, and commercial prospecting.
Its retention section keeps personal data for as long as needed to answer a query or until a consent period ends. None of the five limbs the A band asks for is addressed for platform data: no retention period for detections, evidence, seller records or the IP documents held in Document Management; no deletion commitment on termination; no access control description; no subprocessor list, despite operations in the United States, Spain and mainland China and an AWS deployment stated on the vendor's marketplace listing; and no incident or breach notification practice.
The security commitment is that the company takes all reasonable measures and requires trusted third parties to do the same by contract, with the caveat that data sent over the internet travels at the user's own risk. Verified 12 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.
Liability is addressed only through a standard limitation clause that does not reach the exposure this product creates, which is the C band and the R50 shape confirmed on Eudia. The Legal Notice and Terms of Use disclaims responsibility for loss, injury, claim, liability or damage related to use of the site, for errors or omissions in site content, for site downtime and for third-party networks, and it runs an indemnity in the vendor's favour under which the user holds Red Points and its suppliers harmless.
By its own words that instrument governs the site, so under R50 it is described here and not graded as the platform's position. The service agreement is individualised and unpublished, so no indemnity scope, cap, carve-out, warranty or insurance position can be read before signing. What is published on the AWS Marketplace listing is payment terms rather than liability: fees are due in advance, are based on services purchased rather than usage, and are non-cancellable, non-refundable and not subject to set-off.
Nothing located addresses who bears the loss when a detection is wrong, which on an enforcement product is the exposure that matters: a validated false positive removes a lawful seller's listing and revenue. Verified 12 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, named and functionally described, short of documented depth. Published: direct API connections with Meta, Google and Bing described as enabling immediate action and bypassing geo-restrictions; more than 300 removal agreements with platforms and more than 50 global platform enforcement relationships; coverage of more than 5,000 platforms including Amazon, Alibaba, eBay, Taobao, Walmart, Instagram, TikTok, Telegram and the domestic Chinese platforms Douyin, WeChat, Little Red Book and Pinduoduo; a customer login at a dedicated application subdomain; Document Management holding the IP rights the vendor enforces; and procurement through AWS Marketplace.
Depth is the gap: no developer or API documentation was located on any surface, nothing describes what syncs in which direction or what a customer must configure, and the platform relationships are commercial escalation paths rather than documented integrations an implementer could work from. R15 applies in part to the legal-stack limb: the systems this product's work lives in are the enforcement platforms rather than a document management system, but no integration with an IP portfolio, docketing or trademark management system was located either, which is where an IP counsel's own records sit. Verified 12 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 implied and neither the tenancy model nor the region is stated, which is the C band. The only deployment statement located is on the vendor's AWS Marketplace listing, which records the product as software-as-a-service deployed on AWS, and an AWS qualified software mark appears in the site footer. Nothing on any surface states whether tenancy is shared or dedicated, which regions are available, where data is stored as against where it is processed, or what changes between the Starter, Professional and Enterprise tiers.
Two signals fall short of a statement and are recorded rather than credited: the customer login resolves at a host containing an eu-core segment, which implies a European core region but is a hostname rather than a published residency commitment, and the vendor operates offices in New York, Barcelona, Beijing and Salt Lake City with local teams stated for mainland China coverage, which speaks to where people work rather than where data sits.
No data processing addendum, transfer mechanism or standard contractual clauses were located. Verified 12 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.
Marks appear on the site with no scope, no date and no report available, which is the C band. Every page footer carries a TUV certificate image and an AWS qualified software mark, and neither is accompanied by any text naming what standard, scope or period it covers; the certificate image is not machine-readable to this index, so what the TUV mark certifies could not be established from the site. Also in the footer, and creditable only as programme membership rather than security attestation: Google Trusted Copyright Removal Program member, Google Trusted Reporter, and YouTube Certified Multi-Channel Network.
No trust centre exists on any surface located. The Security Policy page, which is the natural home for this material, is a statement of corporate security principles, responsibilities and objectives naming a Security Manager role and management accountability; it names no standard, no control set, no auditor, no attestation and no report, and it addresses the company's own assets and personnel rather than the customer's data.
Nothing was located to obtain a report through, and no independent security attestation naming Red Points was established. Verified 12 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 vendor refers to proprietary models without identifying what sits underneath, which is the C band verbatim. Published across the Our AI page and the product pages: Vision AI proprietary models scanning logo, shape, packaging, text, colour and known-fake images at once; brand-specific models trained on each protected asset; predictive models trained on 2.7 billion data points analysed monthly; a Seller Risk Score computed across more than 17 million seller identities; an AI Copilot; and a decade of accumulated training described as an encoded brand knowledge layer.
Not one model is named, no provider is identified, no architecture beyond ensemble visual matching is described, and no commitment to notify customers when any of it changes was located. The AWS deployment stated on the marketplace listing is infrastructure and is spent on the deployment axis rather than here, per the ground rules against one fact working two axes. Verified 12 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.
Real pricing is published for part of the range, which is B. The vendor's own AWS Marketplace listing, where Red Points is the seller of record and the listing copy and payment terms are its own, publishes a twelve-month contract dimension: a Professional Program covering three trademarks with marketplace and social media coverage and image protection at $39,000. Under R117(2) that is a vendor-authored storefront and a published price.
The vendor's own pricing page then publishes the structure in unusual depth without any figure: five programmes for counterfeits, gray market, brand impersonation, domain abuse and piracy, each with named tiers, and a feature-by-feature comparison across Starter, Professional and Enterprise covering trademark counts, channel coverage, managed service inclusion, dashboards and the ten-plus add-ons, each cell marked as included, add-on or not available.
The charging model is explained: a flat fee scoped to channels, markets, brand count and reporting depth, with unlimited detections, takedowns and users inside the agreed scope and no per-detection or per-takedown metering. What holds this off A is that every tier on the vendor's own site routes to Request Pricing, so the single marketplace figure is the only number published and it is not the entry rate, since the Starter tier sits below it unpriced. A pricing row is owed under R17 and is written. Verified 12 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.
Coverage is described with substance and the boundaries are only partly stated, which is B. Published precisely: five programme families; channel coverage named platform by platform across marketplaces, social media, search engines, ads, mobile apps, domains, video platforms and AI shopping surfaces; regional coverage named as North America, EMEA, APAC, LATAM and mainland China, with the Chinese platforms listed individually; and buyer functions named on the pricing page as legal, brand protection, e-commerce, security and executive teams, with IP titles appearing among the named customer references.
Size segmentation is stated and so is one real boundary: the pricing material states that Red Points works with mid-market and enterprise companies, which tells a smaller brand it is not the intended buyer. What is left open is everything else about the limits: no statement of which infringement types or jurisdictions fall outside scope, no statement of what the platform does not detect, and nothing on the practice-area dimension.
R15 applies to the firm-segment and practice-area limbs, which do not bite on a product bought by a brand owner's IP and brand protection function rather than by a law firm or a practice group. Verified 12 September 2026.
4 public documents
The public pages on file for Red Points, 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.
-
redpoints.com/our-ai5 signals
Primary Law Corpus Provenance, Good Law Verification, Refusal and Uncertainty Behaviour and 2 more
Read Sep 12, 2026
-
redpoints.com/data-protection-policy3 signals
Prompt and Output Retention, Third Party Request and Subpoena Notice, Outside Counsel Guideline Readiness
Read Sep 12, 2026
-
Client Data in Training, Bar Guidance Alignment
Read Sep 12, 2026
-
redpoints.com/pricing1 signal
Ethical Walls and Matter Segregation
Read Sep 12, 2026
$39,000 per yearUSD, as published, never converted
- Red Points charges one flat yearly fee for an agreed scope of work, and inside that scope it finds and removes as many fakes as it takes.
- The one price published anywhere is on its AWS Marketplace page: $39,000 a year for the Professional Program covering three trademarks, marketplace and social media coverage, and image protection.
- This is the marketplace price for one specific package, not the cheapest plan; the Starter plan costs less and the price is not published.
- What you pay depends on how many brands, channels, markets and reports you need, not on how many fakes are found.
- Fees are paid up front and are not refundable.
One figure is published and the structure around it is published in unusual depth. The figure sits on the vendor's AWS Marketplace listing as a twelve-month contract dimension: Professional Program, covering three trademarks with marketplace and social media coverage and image protection, at $39,000. It is recorded here as the lowest published rate rather than as a floor, because the Starter tier sits below it on the vendor's own pricing page with no figure attached.
The model is a flat fee scoped up front, with unlimited detections, takedowns and users inside the agreed scope and no charge per detection or per takedown, so cost does not rise as infringement volume rises. The vendor names the scope variables that set the fee: channels included, markets covered, number of brands, and the level of seller intelligence and reporting. Five programmes are priced separately, for counterfeits, gray market, brand impersonation, domain abuse and piracy, each with named tiers, and the counterfeits and gray market programmes carry a full feature-by-feature comparison across Starter, Professional and Enterprise marking each line as included, an add-on or unavailable, covering trademark counts, channel coverage, China coverage, managed service inclusion, dashboard counts and more than ten add-ons including investigation services, test purchases, UDRP and trademark registration.
Every tier on the vendor's own site routes to Request Pricing. Payment terms published on the marketplace listing: fees due in advance, based on services purchased rather than usage, non-cancellable, non-refundable and not subject to set-off.
Confidentiality and data terms: No customer agreement, data processing addendum or subprocessor list is published on any surface located. The Legal Notice and Terms of Use states that any provision of services is formalised individually through channels outside the site, so the confidentiality and data terms are negotiated rather than published, and the only contractual language available to a buyer in advance is the payment and refund wording reproduced on the vendor's AWS Marketplace listing.
Note: Figure read from the vendor's AWS Marketplace listing on 12 September 2026, where Red Points is the seller of record and the listing copy and payment terms are its own, which is the R117(2) test. Structure read from the vendor's own pricing page, modified 2 September 2026. Two competitor comparison pages surfaced during this research carrying pricing assertions about Red Points and were excluded outright under the ground rules; no figure from either appears in this row or anywhere on this record.
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?
No customer agreement, terms of service or equivalent contract is published on any surface located, and no policy page states a position on training. Nothing is granted and nothing is withheld, so a client has no term to hold the firm to. Where a policy page does state a position, the row takes the matching policy value instead and the summary records that no agreement exists.
No customer agreement is published on any surface located, and no policy page states a position on training either way. The Legal Notice and Terms of Use is a website instrument by its own terms and says expressly that any provision of services is formalised individually, at the user's request, through channels external to the site, so the terms that would answer this question are negotiated and unpublished. The Data Protection Policy, last updated July 2019, is a website and marketing policy: its collection channels are contact forms, the newsletter, the blog, cookies, demo requests, corporate email, social profiles and the careers page, and its purposes are answering queries, sending news and commercial prospecting.
It states that data will not be used for purposes beyond those in the policy, which is a commitment about website personal data and not about brand assets, product images, trademark records or detection data held in the platform. The Security Policy names no data use. The AWS Marketplace listing reproduces payment terms referring to a Scope of Services, which evidences that a written agreement exists without publishing it. Nothing is granted and nothing is withheld, so a client has no term to hold the brand owner to.
Prompt and Output Retention
How long does the product keep what a lawyer typed, and can that be set to zero?
No located public material states how long prompts and outputs are retained.
No located public material states how long the platform keeps what a customer puts into it or what it generates. The only retention language found is in the website Data Protection Policy and is scoped to enquiry and marketing data: personal data kept for the time needed to deal with a query, or until the end of a consent period, with some retained for legal, regulatory, statistical or historical purposes and anonymised when no longer needed.
Nothing addresses the material that matters here, which on this product is the brand's uploaded trademark records and catalogue images, the detection and evidence record, the Seller Risk Score history and the enforcement audit trail. Document Management is published as one place to store and access the IP rights the vendor enforces, so storage is a named feature, but no period, no customer control and no zero-retention option is stated for it, and no deletion commitment on termination was located.
Ethical Walls and Matter Segregation
Does retrieval respect the firm’s ethical walls, or can the model read across them?
No located public material addresses walls or matter level segregation.
No located public material addresses segregation between users, brands or matters. What is published bears on the question without answering it: the pricing tiers include unlimited users within an agreed scope, live and custom dashboards, and Document Management as a shared store of the customer's IP rights, and the Enterprise tiers add five to ten custom dashboards, so a multi-brand or multi-entity customer is plainly contemplated.
Nothing states whether one brand's detections, evidence or seller intelligence are separated from another's inside a customer account, whether roles or permissions exist, or how access is enforced. The Security Policy speaks to the company's own personnel and assets rather than to a product permission model, and no agreement is published in which a segregation term could sit. Recorded as of 12 September 2026 on the surfaces named in the build log.
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.
Nothing located addresses third party requests for customer data. The closest published statement is in the website Data Protection Policy, which says Red Points will not provide users' personal data to third parties and that if it decides to do so the user will be informed in advance and consent requested. That is a commitment about website and marketing personal data, not about a brand's material in the platform, and it addresses voluntary sharing rather than compelled disclosure: no surface read mentions a subpoena, a court order, law enforcement or any other legal process, and there is no transparency report.
The policy does state that data may be retained to administer the company's rights, including to assert claims in court. With no customer agreement published, the confidentiality section that would ordinarily carry a notice commitment could not be read. Recorded as of 12 September 2026.
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 primary law corpus is identified because the product does not answer from one. What it answers from is published in outline: 2.7 billion data points analysed monthly, more than 17 million seller identities, cross-platform behavioural signals, investigation evidence, live trend capture, and the customer's own trademark and design registrations held in Document Management. The nearest thing to encoded law is the statement that generated evidence is formatted to each infringement type, platform and jurisdiction, which implies jurisdiction-specific requirements are encoded somewhere, and nothing published identifies where those requirements come from, how they are maintained or on what basis the underlying detection corpus is collected or licensed.
The value records that the corpus behind the product's determinations is not identified, which is the honest reading for a product whose source material is marketplace and seller data rather than case law.
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.
The product cites no legal authority, so nothing located addresses checking subsequent history, and nothing would be expected to. The currency mechanisms it does publish are about threats rather than authority: continuous rescreening as new sellers and listings appear, a quarterly discovery audit of infringing platforms beyond a plan's scope, and live capture of new evasion trends. Recorded so the row states the position rather than leaving a reader to infer it from silence.
Refusal and Uncertainty Behaviour
What does the product do when the answer is not in the corpus?
The vendor describes refusal or abstention behaviour in public materials.
The vendor describes an abstention path in public materials, short of anything demonstrable. The Our AI page states that the autonomous agents run the high-volume work and escalate to a person only when it is required, and separates that agentic tier from a governed tier reserved for the decisions the vendor says cannot go wrong, being validation, enforcement and evidence generation. The vendor's own brand protection guide adds that human experts review complex edge cases to prevent false positives and protect authorised sellers.
So the published behaviour when the system is unsure is to hand the case to a person rather than to decide it. What is not published is any threshold, any rate at which escalation occurs, or any evaluation demonstrating it, and the same material claims elsewhere that the AI eliminates false positives, which is an accuracy assertion rather than an abstention description.
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 on 12 September 2026, on both the product name and the company name, against published trackers of AI hallucination decisions including coverage of the Charlotin AI Hallucination Cases database, for court records addressing fabricated or hallucinated legal citations in this product's output. None located. This is a statement about the public record on that one subject as of that date, and this signal does not record enforcement or takedown litigation.
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 engages with bar or ethics guidance. The vendor addresses professional responsibility in one direction only, and firmly: its Legal Notice states that it is not a law firm, that its employees will not act as the customer's attorney, that it does not provide legal advice and does not intend to engage in the practice of law, and that it provides only self-help services at the customer's specific direction.
That is a disclaimer of legal practice rather than engagement with the guidance a lawyer buyer is bound by, and nothing located addresses ABA Formal Opinion 512, any state bar opinion on generative AI, or the duties of a supervising attorney whose enforcement programme runs on this platform. Recorded as of 12 September 2026.
Billing and Fee Posture
Does the vendor address what happens to the bill when the work takes an hour instead of six?
The product does not touch a fee between a lawyer and a client. It operates before an engagement exists, or it is bought by a team that bills no client for the work. Savings claims aimed at the buyer’s own cost are recorded in the summary and do not make the row a savings claim, because no client bill is in the loop.
The product does not touch a fee between a lawyer and a client. It is bought by a brand owner's IP, brand protection or legal function to enforce that company's own rights, and no client is billed for the work the detection and enforcement engine performs. Savings claims are published and are aimed squarely at the buyer's own cost rather than at a client invoice: 10.3 times lower average cost per enforcement against analyst-based models, 12.6 full-time-equivalents of average capacity per brand protected, 130+ hours reclaimed in a quarter at a named customer, and a flat fee explicitly sold as removing analyst hours from the bill.
Under the value's own terms those are recorded here and do not make this a savings-claims row. Two adjacent fee arrangements belong on the record without changing the value: the Revenue Recovery Program is described as zero-cost litigation that recovers funds from infringing sellers, reported at more than $20M to date, and IP registration, UDRP and domain dispute resolution are sold as add-on services, so legal work is billed alongside the platform but not as AI-assisted time charged to a client.
Outside Counsel Guideline Readiness
Can a firm get this vendor through a client’s AI clause without a bespoke negotiation?
No located public material supports a client side disclosure obligation.
No located public material supports a client side disclosure obligation. No subprocessor list, no model provider statement, no data processing addendum and no consent or notification pack was found on any surface read, and no customer agreement is published from which an annex could be forwarded. The Data Protection Policy states only that Red Points requires trusted third parties handling personal data to protect it by contract, without naming any of them, and the vendor operates in the United States, Spain and mainland China with an AWS deployment stated on its marketplace listing, none of which is reconciled into a processor list.
A firm asked by its client to evidence this vendor under an AI or data clause would have nothing published to forward. Recorded as of 12 September 2026.
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.
Elements of a record are published and they are built for platform enforcement rather than for a court's AI disclosure. The Our AI page states that every decision is logged with what was detected, why, what triggered it and what happened next, that legal-ready evidence is generated automatically and formatted to each infringement type, platform and jurisdiction, and that it is filed simultaneously to multiple agents; Vision AI is stated to show the exact logos, patterns or products that triggered a match; and Document Management holds the IP rights being enforced.
That is a per-incident, exportable, jurisdiction-formatted evidence record, which is more than most products in this corpus produce. What is absent is the part a certifying lawyer would need: nothing states that the record identifies which model or classifier made the call, what it retrieved, or which named human verified it before the report was filed, and no disclosure template or court-facing guidance was located.