Ketch
Ketch is a data privacy management platform for privacy, legal and security teams, covering consent management, data subject request automation, data mapping, privacy risk assessments, preference and identity management and AI governance, with a server-side permission store that carries a person's consent and preferences into every connected system. Its AI layer, the Ketch Agent Network, connects to a customer's systems and contracts, classifies data fields by sensitivity and purpose, reads data processing agreements and sub-processor lists to extract what each vendor may do with data, cites the relevant clause, flags where a contract and a system configuration disagree, and keeps the processing record current; a regulations hub synthesises laws and enforcement actions and maps them to where a customer operates. The company is Ketch Kloud, Inc., a Delaware corporation based in San Francisco, and sells a free tier alongside paid plans whose Pro tier adds single-tenant data residency. It publishes its master services agreement and data processing addendum in full, with a contractual commitment not to use customer data to train models, prior notice of compelled disclosure, a government-access clause that commits to challenging unlawful requests, and a thirty-day post-termination deletion; it holds ISO 27001 and SOC 2 Type II and warrants to maintain them for the term. Named industry solutions span retail, technology, financial services, healthcare, media, automotive, travel and education, and the platform is hosted on Amazon Web Services in a multi-tenant environment.
Capability grades
All 15 axes, graded from public sources on the date shown. Hover a grade to see what the letter means on that axis.
AI Centrality
How much of the product is actually AI. Whether the machine learning is the mechanism the buyer is paying for or a feature layered onto conventional software, and whether the vendor is specific about which is which.
The models are the engine of a core capability layered on a product that would still function without them. The Ketch Agent Network is described as the orchestration layer powering every product, with agents dispatched for discovery, risk, consent configuration, documentation and reporting, reading contracts and classifying data fields; an AI Governance module is sold alongside. Remove the models and consent capture and enforcement, the permission store, DSR intake and fulfilment, identity resolution and the regulations hub remain, which is a working privacy platform, and the master services agreement defines AI Technology as one component of the Services rather than the Services themselves. The May 2026 Agent Network release is the point at which the AI layer became central to the marketing; the structure is a privacy platform with a model layer. Home page, platform overview, agent network page and terms 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.
Grounding is real and documented, short of any accuracy figure. The agent network page states that the agent populates answers from what it knows about the customer's systems, configurations and processing activities, that every answer comes with a source, and that what a vendor may do with data is extracted from the legal text and cited to the clause; the regulations hub describes laws, enforcement actions and settlement terms synthesised and mapped to where the customer operates. Master services agreement section 7.3 states that outputs may be inaccurate, incomplete or inappropriate and that the customer must review and validate them before use. No accuracy measurement, test set or evaluation is published. The primary-authority limbs apply only in part: the product cites the customer's own contracts and configurations, and the regulatory content is described as synthesised without naming its sources. Agent network page, platform overview, regulations hub and terms 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.
A written commitment that the models work alongside a reviewing customer, with real review surfaces, short of the full control structure. Terms section 7.3 places responsibility for reviewing and validating AI outputs on the customer before they are relied on; the agent network page states that the agent flags where a contract's permissions and a system's configuration disagree rather than acting on the mismatch, and that every answer carries a source for review; DSR automation carries an auditable trail for every request and the Permission Vault keeps consent decisions as auditable evidence. What is not published is the control structure the A band asks for: no thresholds for when an agent acts alone, no description of which agent actions execute without approval, and no stated route back after an output is wrong beyond the customer's own correction. The AI Governance product page was not opened. Terms, agent network page and platform overview 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.
On the surfaces read, testimonials stand without attribution and the only figure is aggregate scale. The home page carries quotes from legal, marketing and privacy operations teams, including that Ketch is software lawyers can use, without a named person or organisation in the text read, and the site-wide figure of 67.2 billion consent transactions a month is a scale claim rather than a deployment outcome. A Customers page and customer stories exist in the navigation and were not opened, so this grade rests on the home page and the legal documents and is rebuttable on that page. Home page, terms, DPA and platform pages 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.
Four of the five limbs are in the published agreement and the privilege limb is absent. No training: master services agreement section 7.2 provides that Ketch will not, and will not permit any third party to, use customer data to train AI or machine learning models, with Feedback and Usage Data carved out for improving its AI Technology. Segregation at the level an in-house buyer requires: the DPA states a multi-tenant environment with role-based access, SSO and SCIM on the home page, and the Pro plan adds single-tenant data residency. Retention and deletion: DPA Schedule 3 sets retention for the life of the contract plus thirty days followed by automated deletion or anonymisation, and DPA section 9 gives destruction or return on written request. Third-party model providers: section 7.2 binds them on training, the DPA commits sub-processors to the government-access protections of the standard contractual clauses, and a current sub-processor list with countries is published at a URL the DPA names, which could not be opened on this channel. Nothing on any surface addresses privilege or work product handling. Terms, DPA and home page read 6 September 2026.
UPL and Professional Responsibility Posture
Whether the vendor is clear that it supplies a tool rather than legal advice, who its audience is, and how it addresses unauthorized practice of law, competence and supervision duties, and jurisdiction limits. ABA Formal Opinion 512 is the reference point.
A real position on advice versus tooling, short of the supervision dimension framed for a lawyer and of jurisdiction limits. Master services agreement section 3.4 states that the services, documentation and any communication from Ketch are not intended and should not be taken as legal advice, and section 9.2 that outputs are not to be relied on as a substitute for professional advice; section 7.3 requires the customer to validate AI outputs for accuracy, legality and compliance and bars representing them as human-generated. The buyer is stated as privacy, legal and security teams. Nothing addresses how the product supports a supervising lawyer's duties beyond the review obligation, and no jurisdiction limit is named beyond the export-control bar in section 3.3. Terms and home page read 6 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.
A governance commitment without a published mechanism, testing regime or accountable owner. Master services agreement section 7.1 states that Ketch designs and operates its AI Technology with commercially reasonable measures consistent with industry practice and maintains policies and processes reasonably designed to assess and mitigate material risks, which is a contractual undertaking to have a governance process rather than a description of one. An AI Governance product page exists and was not opened; it sells governance of the customer's AI rather than describing Ketch's own. No framework, ISO 42001 or equivalent, pre-release testing description or statement about uneven output is published. Terms and platform navigation 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.
Retention, deletion, access control, sub-processors and incident practice are all published, current and specific enough to hold the vendor to, in the DPA of 7 May 2026 and its Schedule 3. Retention: customer data is kept for the life of the contract and thirty days, then deleted or permanently anonymised by an automated process; server logs are kept twelve months; free services carry no retention obligation. Deletion: DPA section 9 gives destruction or return on written request and terms section 12.5 fixes the thirty-day window. Access control: formal provisioning and deprovisioning, regular access reviews, staff confidentiality agreements and annual training, TLS 1.2 and 1.3 in transit and SSE-S3 encryption at rest, with a named data protection officer. Sub-processors: DPA section 5.2 publishes a current list with countries at a stated URL, with a subscription mechanism for notice of additions and a thirty-day objection right; the page itself could not be opened on this channel and is named as a limit rather than an absence. Incident practice: DPA section 7 commits to notice without undue delay with cause identification and remediation, and Schedule 3 states a six-hour recovery point and twenty-four-hour recovery time objective, annual penetration testing and annual restoration tests. Terms and DPA read in full 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.
What the vendor stands behind is published and specific, including that on wrong output it stands behind nothing, which a buyer can read before signing. Master services agreement section 10.1 gives an IP infringement indemnity with the exclusions in 10.2 and the remedies in 10.3, ending in termination with a pro-rata refund; section 11.1 caps each party's aggregate liability at fees paid in the preceding twelve months and 11.2 excludes indirect loss including loss of data; section 9.1 warrants material conformity to documentation with the exclusive remedies in 12.3 and 12.4; section 7.3 places all responsibility for AI outputs on the customer and 9.2 disclaims warranties of accuracy and reliability; section 13 removes indemnity, warranty and, where enforceable, all liability for free services with a thousand-dollar cap otherwise. No insurance is stated, and order forms and the DPA prevail over the MSA on conflict. Terms read in full 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 some. The agent network page names Braze and Snowflake among hundreds of SaaS connections and describes what the connection does, classifying every field by sensitivity, category and purpose and keeping the inventory current as the stack changes; public documentation exists at docs.ketch.com and an Integrations page exists in the navigation, neither opened. The connections run to data systems, CRMs and marketing platforms rather than to document, matter or e-billing systems, which is what this axis was written for; for a privacy platform sold to an in-house team that is the relevant estate, and the note says so rather than penalising the absence of a DMS. Agent network page, home page and footer 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.
The tenancy model is stated clearly and the residency detail is partial. DPA section 6.2.2 states that Ketch operates a multi-tenant cloud environment, Schedule 3 that the product infrastructure runs on Amazon Web Services across multiple availability zones with backups in a separate geographic region, and the home page that the Pro plan adds single-tenant data residency. No region is enumerated, nothing states where model processing occurs as distinct from storage, and the transfer provisions in DPA section 12 assume third-country transfers from Europe to the United States under standard contractual clauses or the Data Privacy Framework. DPA, terms and home page 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.
Certification is real, stated, contractually maintained and evidenced on request, short of a report reachable without asking. DPA section 6.2.1 states that Ketch holds ISO 27001 certification and SSAE 18 SOC 2 Type II reports for the Services and agrees to maintain them for the duration of the agreement, with reports available on written request subject to confidentiality; Schedule 3 adds ISO 27001:2022, annual independent assessment and annual penetration testing, and the home page adds ISO 27701. A trust centre exists at trust.ketch.com and was not opened, so whether the report is self-serve or sales-gated is not established; under the standing rule for an unstated access tier the lower tier is graded and the trust centre is the rebuttal route. No auditor or coverage period is stated on the surfaces read. Terms, DPA and home page 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.
On the surfaces that could be read, the vendor refers to AI Technology and underlying models without identifying what sits underneath, and the document that would identify it could not be opened. The master services agreement defines AI Technology generically, section 7.4 refers to Ketch's underlying models and algorithms, and the privacy policy names AI and machine learning services as a category of service provider. DPA section 5.2 states that a current sub-processor list with identities and countries is published at ketch.com/subprocessors with a subscription mechanism for notice of additions, which is the change-notification limb met in terms; the page did not surface in search and links inside fetched pages are not fetchable on this channel, so whether it names model providers is not established. This is a limit on this reading and not a finding of non-disclosure; the sub-processor page is the rebuttal route and a provider list on it would lift this to B. Terms, DPA, privacy policy excerpt and home 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.
The unit and structure are stated in the agreement without the figure, and the pricing page was not read. Master services agreement section 5.2 states that fees are based on the tier and usage limits in the order form, measured monthly by the methodology in the documentation, with defined units of Unique Identities and Contacts and automatic movement to the next tier on overage; section 5.1 gives sixty days' notice of renewal price changes, section 12.2 sets automatic renewal for the longer of the initial term or twelve months with thirty days' notice of non-renewal, and section 13 governs a free tier, Ketch Free, offered without warranty. The home page names a Pro tier with single-tenant residency. A pricing page exists in the navigation and was not opened on 6 September 2026; a published figure there would lift this to A and a demo-only page would leave it here. Terms and home 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; the boundaries are left open. Buyers are stated as privacy operations, legal, security and marketing teams, with industry solutions for retail and e-commerce, technology, financial services, healthcare, communications and media, automotive, travel and hospitality, and education, and regulatory coverage stated as GDPR, CCPA and CPRA, emerging US state laws through a US Privacy Law Atlas, and privacy and AI laws generally. What is not stated is where the product stops: no jurisdiction, regulation or industry is named as unsupported, and no law firm use is described. Home page, footer navigation, regulations hub and terms 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?
The published terms prohibit training on customer content. Not a policy page, the agreement.
The commitment is in the agreement. Master services agreement section 7.2 of 7 May 2026 provides that Ketch will not, and will not permit any third party to, use any customer data to train any artificial intelligence or machine learning model. The clause carves out Feedback, which it illustrates as thumbs-up or thumbs-down labelling of AI suggestions, and Usage Data, defined in section 6.3 as log data and metadata, both of which Ketch may use to train and improve its AI Technology; section 6.5 separately permits aggregated de-identified data for reports. Customer data is defined to include outputs produced by the services. 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?
A specific retention period is published and the customer cannot change it.
A specific period is published and the customer cannot change it during the term. DPA Schedule 3 states that customer data, which the agreement defines to include outputs, is retained for the life of the contract and thirty days thereafter, after which an automated process deletes or permanently anonymises it except for what law requires to be kept; terms section 12.5 repeats the thirty-day window. DPA section 9 gives destruction or return on written request after the services end, and server logs are retained twelve months. Nothing states a shorter or configurable window for prompts to the agents during the term, and free services carry no retention obligation at all. 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?
The product maintains its own permission model, documented, requiring the firm to keep it aligned.
The product maintains its own permission model and documents it at the level of a description. DPA section 6.2.2 states that Ketch operates a multi-tenant cloud environment, the home page lists role-based access control, SSO and SCIM as standard, and the Pro plan adds single-tenant data residency; DPA Schedule 3 describes a formal access provisioning lifecycle with regular access reviews. That is tenant-level separation with documented access control, which is what an in-house buyer requires; nothing describes how the agents' retrieval respects role permissions within a tenant. 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?
Terms commit to notice where lawfully permitted. No transparency report located.
The published agreement commits to notice, and goes further than most. Master services agreement section 8.3 requires prior notice of any compelled disclosure of confidential information to the extent legally permitted, with assistance to contest it; DPA section 8.1 requires prompt notice of any legally binding public-authority access request with a summary of its nature, efforts to obtain a waiver where notice is prohibited, a challenge to any request the vendor concludes is unlawful with interim measures sought, disclosure of the minimum permissible, and notice of any direct access by a public authority; Schedule 1 section 1.12 routes notice to the customer rather than the data subject. No transparency report is published. Surfaces checked 6 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?
Coverage is described by jurisdiction with no identification of the underlying corpus.
Coverage is described by jurisdiction without identifying the underlying corpus. The regulations hub and the US Privacy Law Atlas describe laws, enforcement actions and settlement terms synthesised and mapped to where a customer operates, covering GDPR, CCPA, CPRA and US state laws, and the agents draw on the customer's own contracts and configurations with citations to the clause. Nothing states where the regulatory text is sourced, under what licence, or how often it is refreshed beyond the claim that it is synthesised as it lands. Regulations hub, agent network page and home page read 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 cites no case law: its regulatory content is statutes, regulations and enforcement actions, and its agents cite the customer's own contracts. Recorded as the honest value for a product with no 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 agents do when they cannot ground an answer. The agent network page states that every answer carries a source and that the agent flags where a contract's permissions and a system's configuration disagree, which is a mismatch alert rather than an abstention path, and terms section 7.3 places validation on the customer. Nothing describes a confidence signal or a no-answer behaviour. Agent network page, platform overview and terms 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 Ketch or Ketch Kloud, Inc. 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, none of which is this product. This is a statement about the public record, not a finding about the product, and a privacy platform that cites no case law 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 engages with bar or ethics guidance. The master services agreement disclaims legal advice and requires human validation of AI outputs, and the regulations hub engages extensively with privacy statutes and enforcement, but no ethics opinion, bar rule or professional responsibility framework is named on any surface read. Terms, home page and regulations hub 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?
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 buyer is an in-house privacy, legal or security function that bills no client, so the product sits outside a lawyer-to-client fee relationship. The vendor's own charge is a tiered subscription on Unique Identities and Contacts per the master services agreement, and its published savings framing is operational, the processing record that took weeks now maintained continuously; nothing addresses how AI-assisted work is recorded or disclosed on any bill, and no law firm is a named buyer segment. Surfaces 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.
A current sub-processor list is published. DPA section 5.2 states that the list, with identities and countries of location, is maintained at a stated URL with a subscription mechanism for notice before any new sub-processor processes personal data, and section 5.3 gives a thirty-day objection right; the DPA itself is published in full and drafted to be executed by reference. The list page could not be opened on this channel on 6 September 2026, so whether it identifies the model providers behind the AI Technology is not established, which is why the row sits at this value rather than the top one; the page is the rebuttal route.
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 platform keeps auditable trails for data subject requests and consent decisions and the agents cite sources for their answers, but those are records of privacy operations rather than a per-document record of the model used, sources retrieved and human verification, and the product produces no court-facing work product. Agent network page, platform overview and terms checked 6 September 2026.