Stripe Privacy Center (supplemental privacy notices)

Pika · Official agreement

Read the agreement

22 of 266 cards in Inference & Profiling

Inference & Profiling · Company practice

Demandbase may infer a visitor’s company from IP and cookie information

The table lists IP, visit times, referrer, page URL, cookie ID, sign-up data and device/browser characteristics as data made available to Demandbase. Its description says the tag sends IP/cookie information to an identification API to determine the visitor’s company and measure interest. This is the service’s described inference, not a guarantee of correct employment identification. The advertising toggle can be disabled.

Supporting quotations

Demandbase
IP address, timestamps of visit, referrer, page URL, cookie ID, Sign-up information, device and browser characteristics
The Demandbase tag reads the IP address and cookie of each visitor and passes that information to our identification API, which determines which company the visitor works for.
You can disable the toggle in the “advertising” section of our cookie settings page at any time.

Qualifications & exceptions

  • You can disable the toggle in the “advertising” section of our [cookie settings](https://stripe.com/cookie-settings) page at any time.
Also annotated in the same passage (5)

Inference & Profiling · Company practice

Personal data trains fraud and loss-prevention models

Some data collected in providing services, including personal data, is used to improve products, train fraud and loss-prevention models, and analyze product performance as permitted by applicable law and agreements. Radar and Identity models use personal data, including comparisons between transaction data and past data, to recognize potential fraud, impersonation or transactions unlikely to complete. This describes model training, not a claim that all personal data is always used or that scores measure a person’s creditworthiness.

Supporting quotations

Stripe uses some of the data it collects to improve its products and Services, including by training the models it employs for fraud and loss prevention and to analyze the performance of Stripe’s products as permitted by applicable law and agreements.
Personal Data is required to train Stripe’s fraud and loss prevention models, including those employed by Stripe Radar and Stripe Identity.
For example, they compare Personal Data presented in a specific transaction to Personal Data collected in the past to identify when a fraudster is attempting to perpetrate fraud, including to impersonate a Stripe User or their End Customers.

Qualifications & exceptions

  • as permitted by applicable law and agreements.

Inference & Profiling · Company practice

Payment predictions may select, recommend or retry a pathway

Signals are used to calculate how likely a routing, authentication or authorization approach is to succeed. Depending on the service, the prediction may select, recommend or retry a payment pathway. These models aim to improve performance and prevent fraud or loss, rather than produce a standalone assessment of character or creditworthiness. Where automated decisions have legal or similarly significant effects, or law otherwise requires, the publisher provides legally required notices, explanations and choices, subject to applicable legal exceptions.

Supporting quotations

Stripe uses these signals to identify patterns and calculate the likelihood that a particular payment-routing, authentication, or authorization approach will succeed. Depending on the Stripe Service, Stripe may use the resulting prediction to select, recommend, or retry a payment pathway.
These models are intended to improve payment performance and prevent fraud or financial loss. They are not intended to produce a standalone assessment of an individual’s character or creditworthiness.
Where automated decision-making or profiling produces legal or similarly significant effects concerning you, or where otherwise required by applicable law, Stripe will provide the notices, explanations, and choices required by that law.
These rights may be subject to exceptions under applicable law.

Qualifications & exceptions

  • Depending on the Stripe Service
  • They are not intended to produce a standalone assessment of an individual’s character or creditworthiness.
  • Where automated decision-making or profiling produces legal or similarly significant effects concerning you, or where otherwise required by applicable law, Stripe will provide the notices, explanations, and choices required by that law.
  • These rights may be subject to exceptions under applicable law.
Also annotated in the same passage (8)

Inference & Profiling · Company practice

AI training supports fraud prevention, revenue, experience and service recommendations

The publisher trains AI models for purposes including preventing fraud and other harm, increasing Business Users’ authorization rates and revenue realization, improving service experiences, and identifying additional services that may benefit Business Users. Fraud-model training uses different data depending on the fraud being addressed.

Supporting quotations

We train artificial intelligence models that we deploy for a number of purposes, including to prevent fraud and other harm to Stripe, Users, and others, increase authorization rates and revenue realization for Business Users, provide and improve User experience on other Services, and identify when Business Users may benefit from services they are not currently using.
We train artificial intelligence models to prevent fraud in a number of ways, using different kinds of data appropriate to the kinds of fraud we are trying to prevent.

Inference & Profiling · Company practice

Radar training combines service activity with outside fraud signals

Transaction-fraud models powering Radar are trained to recognize fraud including card testing and card cashing. Training uses individual and business activity on the services plus third-party and public information about compromises and attacks. Data may include names, addresses, IP-derived location, payment methods, advanced fraud-detection data, Financial Connections data, and provider signals that transaction details or devices are compromised, fabricated or linked to fraud. The goal is to recognize fraud patterns at transaction time.

Supporting quotations

Our transaction fraud models which power Stripe Radar and are trained to recognize transaction fraud, including card testing and card cashing.
We train such models using data related to activity that individuals and businesses have conducted on our Services. We also use information from third party and publicly available sources including those related to data compromises and fraud attacks.
The training data may include payments Transaction Data, such as names, addresses, location based on IP addresses, and payment method information, advanced fraud detection data, Financial Connections data, and signals from third party providers regarding whether consumer transaction details and device information may have been compromised, fabricated, or associated with fraud.

Qualifications & exceptions

  • The training data may include

Inference & Profiling · Company practice

Issuing fraud models score risk of unauthorized card transactions

Authorization-fraud models for Issuing use activity on cards issued through that program. Training may include transaction, account-holder, authorized-user and merchant data, such as names, emails, card details, addresses and authorization, decline and dispute histories. The models score the risk of a transaction being attempted without authorization; Issuing users can use those scores to block or allow it. Where automated decisions have legal or similarly significant effects, or law otherwise requires, the publisher provides legally required notices, explanations and choices, subject to applicable legal exceptions.

Supporting quotations

We train these models using data related to activity conducted using cards issued through the Stripe Issuing program.
This training data may include Transaction Data, accountholder data, authorized user data, and merchant information, such as names, email addresses, card details, addresses, and histories of authorizations, declines, and disputes.
Based on these features, our models evaluate the circumstances surrounding a transaction presented for authorization on a Stripe Issuing card to score the risk that a given transaction is attempted without authorization. Issuing users can leverage these scores to determine whether to block or allow transactions attempted on cards issued for them.
Where automated decision-making or profiling produces legal or similarly significant effects concerning you, or where otherwise required by applicable law, Stripe will provide the notices, explanations, and choices required by that law.
These rights may be subject to exceptions under applicable law.

Qualifications & exceptions

  • data related to activity conducted using cards issued through the Stripe Issuing program.
  • Where automated decision-making or profiling produces legal or similarly significant effects concerning you, or where otherwise required by applicable law, Stripe will provide the notices, explanations, and choices required by that law.
  • These rights may be subject to exceptions under applicable law.
Also annotated in the same passage (8)

Inference & Profiling · Company practice

Business and representative details train models for fraud and loss risk

The training description covers Business User and Representative identifiers, birth dates, addresses, bank and device details, business industry and website content, service activity such as payments and disputes, past fraud actions, advanced fraud signals from site access, and signs that details are compromised, fabricated or associated with fraud. Data comes from the services, public sources and third parties to recognize business usage patterns associated with possible fraud or losses. The passage refers to “these models” without giving this bullet a distinct model name.

Supporting quotations

Users may pose a heightened risk of fraud and other losses to Stripe, Users, and other participants in the financial ecosystem.
We train these models using risk signals related to Business User and Representative details, such as names, email addresses, dates of birth, physical addresses, bank account information, and device details, business information, such as industry information and website content, patterns of activity on our Services, including payments, disputes, and past fraud actions, advanced fraud signals collected when a Business User or Representative accesses our sites, and signals indicating that Business User or Representative details may be compromised, fabricated, or associated with fraud.
We collect this data from our Services and publicly available and third-party sources. Training on such data allows our models to recognize usage patterns by Business Users that can be indicative of potential fraud and other behavior that may cause losses.

Inference & Profiling · Company practice

Smart Retries learns when to retry failed subscription payments

To increase authorization and revenue for payments Business Users, Smart Retries trains on past transactions associated with customer card details, names, emails, IP addresses and billing/shipping addresses. Features may include timing and outcomes, card address/BIN/type, device IP/user-agent, amount, subscription history/type and product description, plus third-party IP-derived location. It recommends when a failed subscription payment may best be retried.

Supporting quotations

We train artificial intelligence models to increase authorization rates and revenue realization for Business Users who use our payments products.
We train the Smart Retries models on data from past transactions on Stripe to identify the times at which transactions associated with a customer’s card details, name, email address, IP address, billing address, and shipping address have been successful.
This model may be trained on features indicating the time and outcome of past transactions, card features, such as address, bin, and card type, user device features, such as IP address and user agent, and transaction features, such as the amount, subscription type and history, and product description. We also use location based on IP address as calculated by third-party service providers.
Based on this information, Smart Retries recommends to users when may be the best time to make another attempt at processing a failed subscription payment.

Inference & Profiling · Company practice

Adaptive Acceptance predicts how to resubmit declined charges

Adaptive Acceptance decides whether to reattempt charges rejected by a network or issuer and automatically chooses protocol optimizations and authentication pathways. Training uses past card/customer/merchant authorizations and declines, merchant and charge amount, decline reason, and card BIN, issuer and expiry. Predictions estimate success at different times and under different protocols; the publisher uses them to decide submission methods that maximize its and its users’ profit. Where automated decisions have legal or similarly significant effects, or law otherwise requires, the publisher provides legally required notices, explanations and choices, subject to applicable legal exceptions.

Supporting quotations

We use these models to determine whether to reattempt charges that a card network or issuing bank rejected on the first attempt and to automatically select the optimal protocol optimization and authentication pathway for the transaction.
We train the models using data from past transactions on Stripe, including features related to the history of authorizations and declines we have observed for the card, customer, or merchant involved in the transaction, transaction features such as the merchant, amount of the charge, and the reason it was declined, and card features such as the bin, issuer, and expiry date.
Based on these features, the models calculate the probability that a charge will be successful if submitted at different times and according to different protocols. We use this probability of authorization to determine how to submit the charge to maximize profit for Stripe and its Users.
Where automated decision-making or profiling produces legal or similarly significant effects concerning you, or where otherwise required by applicable law, Stripe will provide the notices, explanations, and choices required by that law.
These rights may be subject to exceptions under applicable law.

Qualifications & exceptions

  • Where automated decision-making or profiling produces legal or similarly significant effects concerning you, or where otherwise required by applicable law, Stripe will provide the notices, explanations, and choices required by that law.
  • These rights may be subject to exceptions under applicable law.
Also annotated in the same passage (8)

Inference & Profiling · Company practice

EEA SEPA models predict failures and chargebacks, not standalone creditworthiness

For SEPA Direct Debit payments in the EEA, machine-learning models analyze transaction and payment signals to predict failure or chargeback likelihood, prevent financial loss and improve payment performance. They do not produce standalone creditworthiness assessments and are not the sole basis for lending decisions.

Supporting quotations

For SEPA Direct Debit payments in the European Economic Area, Stripe uses machine-learning models to predict the likelihood that a payment will fail or result in a chargeback. These models analyze transaction and payment signals to help Stripe and its Business Users prevent financial losses and improve payment performance. They do not produce a standalone assessment of an individual’s creditworthiness and are not used as the sole basis for a lending decision.

Qualifications & exceptions

  • are not used as the sole basis for a lending decision.

Inference & Profiling · Company practice

Identity models learn from verification images with improvement consent

Identity compares biometric features in selfies and photo IDs and looks for fraud indicators, such as obscured faces, modified documents or unfamiliar document formats. Its models are trained using images from verification subjects who consented to having their data used to improve Identity.

Supporting quotations

These models compare biometric features in the selfies with those in the documents and spot features common to fraudulent verification attempts, such as images where the subject’s face is obscured by a mask or the document has been modified or does not match a known format. We train Identity’s models using images from verification subjects who have consented to our using their data to improve Stripe Identity.

Qualifications & exceptions

  • verification subjects who have consented to our using their data to improve Stripe Identity.

Inference & Profiling · Company practice

Instant Bank Payments uses bank details for authorization and model training

Instant Bank Payments models evaluate risks of unauthorized bank-account use or insufficient funds. End Users who sign up agree to provide bank details such as ownership, transaction history and balances, both to decide authorization and to train models to recognize when banks are likely to reject future transactions through the service. Where automated decisions have legal or similarly significant effects, or law otherwise requires, the publisher provides legally required notices, explanations and choices, subject to applicable legal exceptions.

Supporting quotations

Instant Bank Payments, which uses artificial intelligence models to evaluate the risk that someone is attempting to misuse the service by conducting a transaction using a bank account they are not authorized to use or an account with insufficient funds.
End Users who sign up for Instant Bank Payments agree to provide us with details from their bank accounts, such as ownership information, transaction history, and balance details, to determine whether to authorize their transactions and train our models to recognize when banks are likely to reject future transactions attempted through the Instant Bank Payments Service.
Where automated decision-making or profiling produces legal or similarly significant effects concerning you, or where otherwise required by applicable law, Stripe will provide the notices, explanations, and choices required by that law.
These rights may be subject to exceptions under applicable law.

Qualifications & exceptions

  • End Users who sign up for Instant Bank Payments agree to provide us with details from their bank accounts
  • Where automated decision-making or profiling produces legal or similarly significant effects concerning you, or where otherwise required by applicable law, Stripe will provide the notices, explanations, and choices required by that law.
  • These rights may be subject to exceptions under applicable law.
Also annotated in the same passage (8)

Inference & Profiling · Company practice

Models identify Business Users who may benefit from additional services

Models may use platform activity and bank or financial information a Business User supplies to find similarities to users who benefited from products previously. The goal includes identifying candidates for additional services such as US Capital. Information can be personal data where the Business User is an individual; this passage does not promise qualification for a service.

Supporting quotations

We also train models to recognize Business Users that are good candidates to use additional Stripe services, such as Stripe Capital in the U.S.
We may train these models using information related to a Business User’s activity on the Stripe platform and, when the user provides them to us, bank account and other financial information.
The models are trained using such information so that they can recognize similarities between the current Business User and other Business Users who have benefitted from our products in the past. To the extent that Business Users are individuals, such data may constitute Personal Data.

Qualifications & exceptions

  • when the user provides them to us
  • To the extent that Business Users are individuals, such data may constitute Personal Data.

Inference & Profiling · Company practice

Radar risk outputs may inform whether activity is allowed, blocked or reviewed

Radar scores estimate whether a payment method is offered by someone unauthorized, rather than rating the person’s character or creditworthiness. Depending on the purchased service, Business Users may receive scores and set transaction rules. Outputs can include scores, indicators or recommendations, which the publisher, Business Users, financial partners or other authorized recipients may combine with other information to allow, block, authenticate or review activity. Except where expressly stated otherwise, fraud-risk scores are not character or creditworthiness assessments. Where automated decisions have legal or similarly significant effects, or law otherwise requires, the publisher provides legally required notices, explanations and choices, subject to applicable legal exceptions.

Supporting quotations

These scores are designed to identify fraudulent transactions, they do not rate the character or creditworthiness of the individuals involved in a particular transaction.
Based on the service selected by the Business User, Stripe may provide the Radar scores to its Business Users to help them to combat fraud and provide a mechanism for them to set rules to better manage transactions based on Radar scores and other indicators of fraud.
The outputs of these models may include scores, risk indicators, or recommendations. Depending on the service, Stripe, a Business User, a Financial Partner, or another authorized recipient may use those outputs together with other information when deciding whether to allow, block, authenticate, or review an activity. Except where expressly stated otherwise, a fraud-risk score is not an assessment of an individual’s character or creditworthiness.
Where automated decision-making or profiling produces legal or similarly significant effects concerning you, or where otherwise required by applicable law, Stripe will provide the notices, explanations, and choices required by that law.
These rights may be subject to exceptions under applicable law.

Qualifications & exceptions

  • Based on the service selected by the Business User
  • Except where expressly stated otherwise, a fraud-risk score is not an assessment of an individual’s character or creditworthiness.
  • Where automated decision-making or profiling produces legal or similarly significant effects concerning you, or where otherwise required by applicable law, Stripe will provide the notices, explanations, and choices required by that law.
  • These rights may be subject to exceptions under applicable law.

Not stated here

  • The linked locale-specific Radar page is not independently supplied; this card relies on the saved source’s explicit wording rather than additional terms at that link.

Cross-references

Also annotated in the same passage (8)

Inference & Profiling · Company practice

Merchant Risk Tooling supplies Connect platforms with reasoned fraud-risk scores

Merchant Risk Tooling applies internal models to Connected Accounts and gives platform fraud teams reasoned scores of indicators consistent with fraudulent use. The scores supplement platform due diligence and do not rate individuals’ character or creditworthiness. Inputs include payment transactions, Representative IP/physical/email addresses, business website content and other business information.

Supporting quotations

Stripe Merchant Risk Tooling leverages Stripe’s internal risk models to help Stripe Connect Platforms identify indicators of potential fraud associated with their Connected Accounts.
Merchant Risk Tooling produces reasoned scores that identify to Platform’s internal fraud teams when there are indicators consistent with a Connected Account attempting to use the Platform’s services (and Stripe) for fraudulent purposes. These scores are designed to supplement the Platforms’ due diligence related to their Connected Accounts and their business activities. They do not score the character or creditworthiness of the individuals involved in the Connected Accounts’ operations.
The data factored into Merchant Risk Tooling includes payments Transaction Data, business Representative details (such as IP address, physical address, and e-mail address) business website content, and other information regarding the businesses Business Users operate on Stripe.

Qualifications & exceptions

  • These scores are designed to supplement the Platforms’ due diligence related to their Connected Accounts and their business activities.
  • They do not score the character or creditworthiness of the individuals involved in the Connected Accounts’ operations.

Inference & Profiling · Company practice

Radar combines data across services to assess potentially unauthorized payments

Radar uses machine learning and other techniques to detect potential transaction fraud, drawing on data across services. End Customer inputs include payment transactions, advanced fraud-detection data, Bank Connections data, IP addresses and physical addresses. These assess whether the offered payment method is likely unauthorized.

Supporting quotations

To do this, Radar leverages data collected across our Services.
Radar uses data collected about the End Customer from various sources, including payments Transaction Data, advanced fraud detection data, Bank Connections data, IP address, and physical address information. Radar uses this data to assess whether the payment method offered by the End Customer is likely unauthorized.

Not stated here

  • The linked advanced-fraud documentation is not supplied here; no additional collection details are inferred from it.

Inference & Profiling · Company practice

Radar scores can drive automated allow, block or review rules

Radar’s AI score assesses transaction-fraud likelihood. Business Users can use it for automated rules to allow, block or flag transactions for review, and may use Radar as one of several inputs in fraud decisions. This passage describes transaction risk, not a standalone judgment of an individual. Where automated decisions have legal or similarly significant effects, or law otherwise requires, the publisher provides legally required notices, explanations and choices, subject to applicable legal exceptions.

Supporting quotations

Radar’s artificial intelligence model produces transaction “scores” indicating the model’s assessment of the likelihood that a transaction is fraudulent. Business Users can leverage this score and use it to implement automated rules to determine whether to allow, block, or flag transactions for additional review. Business Users can use Radar as one of multiple inputs in making decisions with respect to the potential for fraud in a transaction.
Where automated decision-making or profiling produces legal or similarly significant effects concerning you, or where otherwise required by applicable law, Stripe will provide the notices, explanations, and choices required by that law.
These rights may be subject to exceptions under applicable law.

Qualifications & exceptions

  • Business Users can use Radar as one of multiple inputs in making decisions with respect to the potential for fraud in a transaction.
  • Where automated decision-making or profiling produces legal or similarly significant effects concerning you, or where otherwise required by applicable law, Stripe will provide the notices, explanations, and choices required by that law.
  • These rights may be subject to exceptions under applicable law.

Cross-references

Also annotated in the same passage (7)

Inference & Profiling · Company practice

AI interactions may produce data used for model training and evaluation

Automated/AI interactions may collect contacts, interaction contents, recordings, transcripts and technical or engagement information. Where law permits, recordings and transcripts may be used for staff training, quality assurance, research, operational improvements and training, testing, evaluating and improving AI/ML models. Notice is provided for recorded calls or as legally required. At the start prompt you can opt out of recording, though the conversation may need to move to another channel; the general call section excepts legally required recording from its consent/decline offer.

Supporting quotations

When you interact with an automated or AI-powered communication tool, Stripe may collect and process your contact information, the contents of the interaction, and associated recordings, transcripts, and technical or engagement information.
Stripe may use recordings and transcripts generated through these interactions for staff training, quality assurance, research, operational improvement, and to train, test, evaluate, and improve artificial intelligence and machine-learning models, where permitted by applicable law.
Stripe will provide notice where a call is recorded or where otherwise required by applicable law. If you do not want a call to be recorded, you may opt out when prompted at the beginning of the call. Depending on the reason for the call, opting out of recording may require the conversation to continue through another communication channel.
unless we are legally required to record the call, will ask for your consent to record the call and/or offer you the option to decline the call being recorded.

Qualifications & exceptions

  • where permitted by applicable law.
  • Depending on the reason for the call, opting out of recording may require the conversation to continue through another communication channel.
  • unless we are legally required to record the call
Also annotated in the same passage (1)

Inference & Profiling · Company practice

Personal data may support AI development and improvement where law permits

The publisher may process personal data to develop, train, test, evaluate and improve AI/ML models used to provide, protect and improve its services, where permitted by applicable law. This general statement does not override product-specific consent or purpose limits.

Supporting quotations

Where permitted by applicable law, we may also process Personal Data to develop, train, test, evaluate and improve artificial intelligence and machine-learning models used to provide, protect and improve our Services.

Inference & Profiling · Company practice

Radar and card verification may generate automated risk outputs and share limited data

For End Customers, Radar/card verification detects fraudulent cards and unauthorized logins, may share limited data with requesting merchants for attempted-transaction risk, and may use card images for service improvement. Automated/AI models can generate fraud, failure and chargeback scores or recommendations to authenticate, approve, block or review. Inputs include transaction/contact/payment/card-image and purchase data, phone/tax ID, browsing/usage/referrer/location/cookie/device/IP/physical-address information and model inputs/outputs. The stated basis is legitimate interests in preventing fraud/loss/unauthorized payments and improving risk models. Where law requires, notices and rights are provided for automated decisions with legal or similarly significant effects.

Supporting quotations

In providing such Services, we may provide Business Users that have requested such Services with limited Personal Data about End Customers so that the Business Users can assess the fraud risk associated with an attempted transaction by its End Customer. We may also use payment card images to improve our Business Services. We may use automated tools and artificial intelligence or machine-learning models to assess fraud, payment-failure and chargeback risk and to generate scores, indicators or recommendations concerning authentication, approval, blocking or review.
Transaction information. This includes: name, email address, billing and/or shipping address, payment method information (such as credit or debit card number, bank account information or payment card image), merchant and location, purchase amount, date of purchase, and in some cases, some information about what you have purchased, phone number and tax-related ID.
This includes web browsing information, usage data, referring URLs, location, cookies data, device data and identifiers. Data may also include model inputs and outputs, such as fraud-risk scores, payment-failure or chargeback predictions, and recommendations concerning a transaction or account.
IP address and physical address.
Our legitimate interests in detecting, monitoring and preventing fraud, loss, and unauthorized payment transactions, as well as developing and improving our fraud and payment-risk models. Where required by applicable law, we provide notices and rights relating to automated decisions that produce legal or similarly significant effects.

Qualifications & exceptions

  • Business Users that have requested such Services
  • Where required by applicable law, we provide notices and rights relating to automated decisions that produce legal or similarly significant effects.

Inference & Profiling · Company practice

Capital eligibility assessment uses existing account and financial data

Before a financing offer, the publisher evaluates Business User or Connected Account data. Potential inputs include processing volume, chargeback rate, relationship length, bank balances, transaction history and creditworthiness. This financing assessment is distinct from fraud scores described elsewhere as not evaluating creditworthiness.

Supporting quotations

In Capital for Platforms, we follow the same approach, except that we will review existing data linked to a Connected Account’s Stripe Account to assess their eligibility.
The following information may be considered prior to the offer of financing in order to determine eligibility, including:
Payment processing volume
Chargeback rate
Duration of relationship with Stripe
Bank account balances
Transaction history
Information about your creditworthiness

Cross-references

Inference & Profiling · Company practice

Automated tools may make or support payment decisions without manual review

Uses include services, fraud/financial-loss prevention, payment performance and processes. Examples are risk assessment, apparent funds availability, payout pauses on risk indicators, backup payment choice, authentication/retry/block/review decisions and dispute/chargeback/claim evaluation or applicable resolution. Inputs may include transactions/history, device/browser/account activity, fraud signals, business/partner information and third-party/public sources, depending on service and decision. Significant-effect decisions carry applicable legal notice/choice requirements described in the accompanying rights card.

Supporting quotations

Stripe may use automated tools to process Personal Data and make or support decisions without manual human review.
Examples of automated decisions may include:
assessing fraud or financial-loss risk associated with a payment, payout, Business User, End User, or End Customer;
checking whether sufficient funds appear to be available for a transaction;
pausing a payout where risk indicators are detected;
selecting or recommending a backup payment method;
deciding whether a transaction should be authenticated, retried, blocked, or referred for further review; and
evaluating disputes, chargebacks, or claims, including determining what information is required or, where applicable, resolving a dispute.
The information considered depends on the Stripe Service and the decision being made.
Where automated decision-making or profiling produces legal or similarly significant effects concerning you, or where otherwise required by applicable law, Stripe will provide the notices, explanations, and choices required by that law.

Qualifications & exceptions

  • The information considered depends on the Stripe Service and the decision being made.
Also annotated in the same passage (8)