Legal information
Privacy Policy
How AI Krmítko processes personal data of visitors, customers, users, and customer contacts.
Version 2026-10-05.1. Effective from 2026-10-05.
1. Controller and contact
The controller is Martin Kučera, Družstevní 705/IV, 503 51 Chlumec nad Cidlinou, Czech Republic, Company ID 00959324. For privacy questions or rights requests, use podpora@aikrmitko.cz, the controller’s postal address, +420 606 931 426, or Czech data box yui7vyc. General communication: info@aikrmitko.cz. No data protection officer has been appointed.
2. Processing roles
The Provider acts as controller for visitors to its own public website, accounts, orders, licences, payments, and its own security. For customer-site personal data fetched or supplied for a customer-ordered audit, content processing, or related support, it acts as processor under the DPA and documented instructions. Public availability alone does not remove personal data protection. A customer that itself acts as a processor needs the relevant authorisation from its controller to engage the Provider.
3. Data categories
- Identity, contact, billing, and account data, such as name, email, organisation, locale, and role.
- Order, payment, subscription, invoice, licence, entitlement, and service usage data. Stripe, not the Provider, processes full card numbers.
- Connected-site data, including URL, domain, WordPress, PHP and plugin versions, service status, audit findings, and publicly fetched content.
- Technical and security data such as an IP hash where the relevant application protection is applied, user-agent, time, URL, event type, diagnostics, and audit trail. Infrastructure and server HTTP logs may also process the full connection IP address; not all operational logs use hashing.
- When reCAPTCHA v3 protection is enabled, technical browser and interaction data from login or registration submitted directly to Google for abuse assessment.
- Customer communications, support requests, and information voluntarily supplied in a form or message.
- Agency-report branding settings, including a display name, contact line, colour, and an optional logo converted before storage to a bounded metadata-free PNG.
- Private WordPress content only after explicit opt-in for a feature that requires it.
4. Purposes and legal bases
- Entering and performing a contract with the individual concerned: accounts, licences, sites, subscriptions, support, and ordered features. Contacts and access for people representing an organisation are processed on the legitimate interest in managing the contractual relationship and communications; customer-content processing follows its documented instructions and the DPA.
- Legal obligations: accounting, tax, consumer requests, and cooperation with authorised public bodies.
- Legitimate interests: security, fraud and abuse prevention, operational diagnostics, legal claims, and proportionate service improvement.
- Consent: only where required, such as optional marketing. Consent can be withdrawn for the future. Product opt-in for private-content transmission is an instruction and permission to use a feature, not itself consent from everyone whose personal data appears in that content; the customer must establish its own lawful basis.
- Data needed for an account, entering and performing a contract, or legal duties must be provided; otherwise the relevant service or order cannot be handled. Optional data and marketing consent are not conditions of the basic service. Customer content and contacts may come from the customer or its public website.
5. Recipients and providers
- Providers operating under the WebSupport brand supply hosting, infrastructure, related backups, and system email under standard terms including data-processing provisions. They act as processors of the Provider’s own operational data and as subprocessors of entrusted customer content. The standard Czech terms name ACTIVE 24, s.r.o. and its subprocessor WebSupport s.r.o.; this does not assert an individual audit of every provider contract or that all copies are located exclusively in the EU.
- Google for reCAPTCHA v3: when login and registration protection is enabled, it processes Customer Data as processor under the Google Cloud Data Processing Addendum; the Provider is controller. Since 2 April 2026, Google does not act as an independent controller of that data. The script loads after form submission and sends device, browser, IP address, and interaction data for security, fraud, and abuse prevention. The processing framework consists of the standard Google Cloud Terms and Cloud Data Processing Addendum applicable to reCAPTCHA; this does not assert a separately negotiated individual agreement or determine Google’s roles for other products.
- Stripe: the prepared payment integration for future approved sales. It acts as processor when executing payments on the Provider’s instructions and as independent controller for its own legal obligations and certain security and fraud-prevention purposes. The specific entities and services must be verified against the payment account and contract. Public checkout is currently unavailable.
- Bank transfers are not currently offered, and no bank is named for them. The actual data recipient and its role will be added before any introduction.
- Legal, accounting, security, and technical advisers only as necessary and subject to appropriate confidentiality.
- Public authorities where disclosure is required by law or a binding order.
6. AI and international transfers
The current production configuration does not use OpenAI, Anthropic, Gemini, or another external AI API. The internal aiv-template-v1 provider does not send inputs outside the service infrastructure.
Google and, where applicable, Stripe or their subprocessors may process data outside the EU/EEA; reCAPTCHA cannot be assumed to process data only in the EU. The Google Cloud Data Processing Addendum and Stripe documents govern international transfers, including Standard Contractual Clauses and, where applicable, valid adequacy decisions for specific recipients. Under Google’s current published framework, transfers to Google LLC in the USA and its onward transfers use the EU–US Data Privacy Framework; other restricted transfers use the applicable Standard Contractual Clauses under the Cloud Data Processing Addendum. The former Privacy Shield is not a valid safeguard and is not relied upon. For a new provider or changed transfer, the applicable valid mechanism and necessary safeguards will be assessed before processing begins. Information about the applicable safeguards and how to obtain a copy may be requested at podpora@aikrmitko.cz.
7. Retention
- Raw hosted-generation input: 0 days and discarded after completion. Temporary asynchronous input is encrypted only until a job completes or fails.
- Stripe/provider webhook payloads and queue diagnostics: 30 days.
- Server HTTP logs: daily rotation retaining 14 rotated files; their actual age depends on successful rotation. The application log currently uses a single file without automatic age expiry; it is retained only as necessary for operations, security, incidents, and documented legal claims. Necessity is reviewed manually at least monthly, and unnecessary records are deleted or anonymised after verification; documented legal holds are exceptions. No fixed automatic day-based expiry is promised. The 30-day queue-diagnostic retention does not automatically apply to all these logs.
- WordPress plugin crawler logs: automatic 30-day retention with scheduled pruning.
- SaaS crawler events: 90 days.
- Signed site-heartbeat history: 30 days; the newest evidence for each site remains as an explicit continuity exception. Site-sync idempotency digest and signed result: 8 days by default and never fewer than 7 days.
- Audit runs, usage events, LLM requests including the encrypted full generated result, licence audit logs, and account-security audit events: a 365-day cutoff for scheduled pruning; the open usage period remains until it closes. Zero raw-input retention does not mean zero generated-result retention.
- Closed support requests and messages become eligible for scheduled daily pruning 365 days after closure. Actual deletion may occur during a later successful run depending on the backlog and cleanup availability; open requests and records under an explicit legal hold are not automatically deleted.
- Organisations, sites, and licences: for the contract and afterwards only as needed for legal claims or obligations; the verified organisation-erasure process deletes associated operational data or anonymises links unless a legal reason requires retention. This process alone does not delete the personal user account, its name, email, authentication data, sessions, or password-reset records. Personal-account erasure requests must be verified and handled separately; organisation erasure must not be represented as complete erasure of an individual.
- Agency-report PDF and CSV files are streamed without retaining the generated file. Stored branding settings and the logo are removed during organisation erasure.
- Accounting and tax records: for the period required by applicable law. Our database backups run daily; after a successful backup, files approximately older than 15 days are removed. An unsuccessful run can extend retention until the next successful pruning. Separately, one latest verified deployment rollback backup is retained; it may contain older operational data and has no fixed time expiry, being replaced after the next verified deployment. Provider backups and any infrastructure snapshots follow the retention and erasure terms of the relevant hosting service; our database period does not automatically apply to them. Backup access is restricted, and copies serve recovery, security, and legitimate legal obligations, not new use of customer content.
8. Individual rights
Subject to applicable law, you may request access, correction, erasure, restriction, and portability; object to legitimate-interest processing; and withdraw consent. Send requests to podpora@aikrmitko.cz, the controller’s postal address, or Czech data box yui7vyc. The requester’s identity and authority may be reasonably verified first.
You may complain to the Czech Office for Personal Data Protection, Pplk. Sochora 27, 170 00 Prague 7, uoou.gov.cz, or another competent EU/EEA supervisory authority.
9. Cookies, profiling, and security
The public site uses no analytics or marketing cookies, and fonts are self-hosted. The SaaS uses the session and security technologies described in the Cookie Policy; theme preference is stored locally in the browser. The Cookie Policy explains details and the conditions for necessity.
When reCAPTCHA v3 is enabled, submitting a login or registration form loads Google resources and the _grecaptcha security cookie or related abuse-assessment technologies. Google processes reCAPTCHA Customer Data on the Provider’s instructions under the Google Cloud Data Processing Addendum for security purposes, not personalised advertising. Requests to Google’s domain may also send Google cookies already set in the user’s browser; use cannot therefore be guaranteed to involve only one new security cookie.
The service makes no decision based solely on automated processing that produces legal or similarly significant effects for a user. Safeguards include access control, encrypted transport, hashed tokens and IP addresses, tenant isolation, audit logs, limited retention, and encryption of temporary payloads.
10. Changes
This Policy will be updated when processing changes, particularly before enabling an external AI API or a new payment or analytics provider. The date and material changes will be published at the service’s public address.