Who Trains on Your Data

This table quotes what each company's own terms and privacy pages say about training on customer data, as of the date checked.

Checked

Select lanes and positions. If none are selected, every row stays.

Lane

Position

Showing 20 of 20

Sort

Who Trains on Your Data
ProductQuote
AmpAmpCoding agentsSays customer data is used for training only if the customer opts in or agrees
Amp does not use User Content to train AI models unless You have expressly opted in to such use through the Services.

Terms - Amp(d) Model Training

NotesSame terms: "training cannot be enabled for Enterprise workspaces." Terms name Amp Frontier Corporation.

Browser UseBrowser Use (browser agents)Agent toolsSays it may use customer data or inputs to train its models
We use Inputs to train our AI models in order to improve the Services.

Privacy Policy | Browser UseTo develop and improve our Services

NotesPrivacy policy, under "To develop and improve our Services". Names Browser Use, Inc.

ClericCleric (AI SRE)Agent toolsSays customer data is not used to train AI models
Your data is never used to train or fine-tune models

Security & Data Privacy | Cleric DocsModel & Inference Privacy

NotesPrivacy policy names Agentik, Inc. Terms of use and privacy policy have no training clause.

CodeRabbitCodeRabbit (AI code review)Coding agentsSays customer data is not used to train AI models
Neither CodeRabbit nor its Third-Party AI Model Providers use Customer's code to train any AI models.

Terms of Service | CodeRabbit3.3 AI

NotesThe clause says "Customer's code". The privacy policy separately says "We use OSS to train our systems." Terms name CodeRabbit, Inc.

CognitionDevinCoding agentsSays it may train on customer data, with an opt-out on some plans
Cognition may use Customer Data for model training purposes and to improve and enhance the Services. If you subscribe to a paid Service Tier, you may opt out of this use (“Opt-Out”).

Platform Terms of Service | Cognition3.3.1 Use of Customer Data

NotesDevin security docs say: "If you are an Enterprise customer, we will never train on your data without your express prior written consent." Terms name Cognition AI, Inc.

FazeshiftFazeshift (accounts receivable agents)Finance back officeSays customer data is not used to train AI models
Your AR data does not train anyone’s model.

Security & Compliance | SOC 1 & SOC 2 Type II | FazeshiftAI and your data

NotesSame page: "Fazeshift runs on models from OpenAI and Anthropic under zero-data-retention agreements." The privacy policy has no training clause; no terms page was found at /terms.

GreptileGreptile (AI code review)Coding agentsSays it may use de-identified, aggregated or derived data from customer use to improve its products or models
Except as otherwise provided in an Order and subject to the other provisions of this Section 2.2, Company may (i) aggregate and anonymize Customer Data ("De-Identified Data") for analytical purposes and to monitor, improve, or expand the Services, Platform or Company's commercial offerings, and/or (ii) use such data to train and improve artificial intelligence algorithms and models ("AI Training and Learnings").

Terms of Service | Greptile2.2 De-Identified Data and AI Training

NotesSame section: "However, if Customer does not desire to permit Company to engage in AI training, an organization admin may open Organization Settings, go to Data & Privacy, and turn off Help us improve Greptile."

GumloopGumloop (AI automation)Agent toolsStates different training terms by plan
We do not use Content from the Enterprise Tier to train AI models unless you ask us to in writing.

Privacy policy | Gumloop21. AI Model Training

NotesSection 21 also says, for the Self-Serve Tier: "We may use Self-Serve Tier Content to develop, train, and improve our AI models." Privacy policy names AgentHub Inc. (doing business as Gumloop).

KinterKinter (accounting close agents)Finance back officeSays it may use de-identified, aggregated or derived data from customer use to improve its products or models
Customer acknowledges and agrees that Kinter may use de-identified, aggregated, anonymized, statistical, or synthetic data derived from Customer's use of the Kinter Platform for any lawful purpose, including for training and improving Kinter's products, services, and AI systems, without compensation or attribution to Customer.

Terms & Conditions - Kinter4.2 Data Collection and AI Training

NotesThe security page says: "Kinter does not use actual customer data for model training." The terms quoted here refer to "de-identified, aggregated, anonymized, statistical, or synthetic data derived from Customer's use of the Kinter Platform". Terms name Kinter Inc.

KodusKodus (AI code review)Coding agentsSays customer data is not used to train third-party foundation models unless enabled
We do not use your customer code or personal data to train or fine tune foundation models (such as models provided by Anthropic, OpenAI, Novita or others), except where you have expressly enabled such a feature in writing or via the Services.

Terms of Use - Kodus6.4 Use of data for AI

NotesThe clause refers to "foundation models (such as models provided by Anthropic, OpenAI, Novita or others)". Terms name Kodus LLC.

LindyLindy (AI agents)Agent toolsSays customer data is not used to train AI models
We do not use your Contributions to train AI models, and our AI model providers have agreed not to use them to train their models.

Terms of Service - Lindy14. CONTRIBUTION LICENSE

NotesTerms section 9, "USER GENERATED CONTRIBUTIONS", defines "Contributions". The security page says: "Lindy and our providers do not use your data to train AI models." Terms name Crivello Corp.

MonkMonk (accounts receivable and collections)Finance back officeSays customer data is not used to train AI models
Your data is not used to train Monk's systems or any model provider's models.

Security | MonkIs my data used to train AI models?

NotesPrivacy policy names Endurance Solutions Inc. Privacy policy and terms have no training clause.

NanonetsTrailAgent toolsSays it may use de-identified, aggregated or derived data from customer use to improve its products or models
Customer hereby agrees and understands that Nanonets may freely use Derived Data for its internal business purposes (including without limitation, for purposes of improving, testing, operating, promoting and marketing Nanonets's products and services). Notwithstanding the foregoing, Derived Data derived from Customer Data obtained via Google Workspace APIs or Microsoft APIs shall not be used to develop, train, or improve generalized artificial intelligence and/or machine learning models.

Terms of Service | Trail6.3 Derived Data

Notestrailhq.com says "Trail is a Nanonets company." The terms at trailhq.com/terms are the Nano Net Technologies Inc. terms, section 6.3, also published at legal.nanonets.com/terms.

ParahelpParahelp (AI support agents)Agent toolsSays customer data is used for training only if the customer opts in or agrees
We never train AI models on your tickets or your customers’ data unless you opt in.

Security | ParahelpDo you use our support tickets to train AI models?

NotesThe same page also says: "We never use your data to train general AI models." The website terms say: "Access to and use of the Parahelp platform, including app.parahelp.com, our AI agents and related services, is governed by the written agreement between the customer and Parahelp (Copilot Labs, Inc. or OWND ApS, as stated in that agreement), together with any order form and Data Processing Addendum."

PromptlessPromptless (docs agent)Agent toolsSays customer data is not used to train general-purpose models; allows use for that customer
User Content and usage data will not be used to develop, train, or improve general-purpose artificial intelligence or machine learning models. However, User Content and usage data may be used to fine-tune or customize artificial intelligence or machine learning models solely for the purpose of providing Services to you.

Terms of Use | Promptless DocumentationMachine Learning and Artificial Intelligence

NotesTerms name Promptless, Inc.

QodoQodo (AI code review)Coding agentsSays it may use customer data or inputs to train its models
Customer hereby grants to Qodo and its Affiliates a non-exclusive, royalty-free, limited (during your engagement with Qodo), worldwide license to use all Customer Data provided to Qodo in connection with these Terms, to (i) permit Qodo to provide the Services to Customer as set forth in these Terms and (ii) to train and improve the Platform’s, the Qodo Models and Services’ performance and to develop new products and services, all subject to Qodo’s compliance with applicable law and privacy regulations.

Terms of Use | Qodo

NotesTerms name Codium Ltd. (dba Qodo). The license is described as "limited (during your engagement with Qodo)".

RavennaRavenna (service desk agents)Agent toolsSays customer data is not used to train AI models
For the sake of clarity, Ravenna Software will not use Customer Data to train, fine-tune, or otherwise develop any artificial intelligence or machine learning models.

Ravenna Software Cloud Services Agreement5.1 Use of Customer Data

NotesCloud Services Agreement section 5.1. Names Ravenna Software, Inc.

RexRex (accounts receivable agents)Finance back officeSays customer data is not used to train AI models
Your Data is not shared across tenants and is not used to train or fine-tune any AI models.

Terms of Service | Rex5. Your Data

NotesTerms name Rex AI Ltd. The security page says: "Your data is never used to train AI models."

SkyvernSkyvern (browser automation agents)Agent toolsSays it may use de-identified, aggregated or derived data from customer use to improve its products or models
We may use anonymized or aggregated data for service improvement and model training;

Terms of Service: Skyvern7.2 Data Processing

NotesNext item in the same list: "You may opt out of model training through an enterprise agreement." Terms name Ikonomos Inc., doing business as Skyvern.

tofutofu (accounting document capture)Finance back officeSays customer data is used for training only if the customer opts in or agrees
7.9 The foregoing license becomes effective only after the Customer has affirmatively opted-in via the in-product checkbox labelled “I agree that my uploaded data may be used to improve Tofu’s AI systems as described in the Privacy Policy and Terms of Service.” The Customer may withdraw this consent at any time in the account settings; such withdrawal will have prospective effect only.

Tofu | Terms & Conditions7.9

Notes"The foregoing license" refers to section 7.8, which grants a licence "solely for the purpose of training, evaluating and improving the Company’s artificial-intelligence systems and related Services". Section 7.10 says: "The Company does not use unmasked personal data contained in Customer Data for AI training. The Company maintains technical and organisational measures designed to ensure that personal data is masked, redacted or de-identified prior to ingestion into any AI training process." Terms name Tofu2 Limited.

Rights are reviewed with counsel before anything moves.

How this index was built

Each row quotes one clause from a company's own public terms of service, privacy policy, data processing agreement, security page or trust documentation, opened and saved on the date shown. The quote is copied exactly as it appears on that page, with no edits. Where a company's marketing pages and its terms or privacy policy differ, the row quotes the terms or privacy policy, and the notes quote the other page. Where training terms differ by plan, the row says so. Each row has a position label that describes what the quoted page says. The labels are not ratings and do not judge any company. Rows were left out when a statement appeared only on a marketing page, when no public terms or privacy policy could be found, when no training statement could be found, or when a recent acquisition left it unclear which terms now apply.

Disclaimer

This index summarizes public pages as they read on the date checked. It is not legal advice, and it does not describe any private agreement between a company and its customers, which may say something different. Companies change their terms, so check the linked page for the current text.

To request a correction, email hello@pillarsdata.com with the company name and the page that changed.