Privacy Policy

1. Introduction

At AutoCISO, we respect your privacy and are committed to protecting it. This Privacy Policy explains how we collect, use, and share information when you use our Service. The Service includes the web dashboard, the public API and SDKs, the MCP server for AI clients, and the AutoCISO Audit Collector browser extension.

2. Information We Collect

Account & Platform Data

Account Data: Name, email address, and company details when you sign up or accept an invitation. If your organisation uses company single sign-on, your account can be created automatically at your first sign-in. In that case we use the name and email address that your organisation's identity provider sends.
Usage Data: Feature activity, audit run counts, AI chat request counts, API usage, and other usage metrics used to enforce subscription limits and to improve the Service.
Technical Data: IP address, browser type, device information, and the client software that calls the API or the MCP server.
API token records: Token name, scopes, the user who created it, expiry date, and last use. When a token reads personal data, the audit trail records the token and the user it belongs to.

Workforce & Access Data

The Service is designed to process your organisation's identity and access information. When you upload or enter data, or send it through the API, we store:

Employee records: Names, email addresses, job titles, departments, managers, locations, phone numbers, employment status, start dates, and end dates of employees and contractors.
Access data: Mappings between employees and SaaS applications, including assigned roles, privilege levels, account state, last-seen activity, dormancy status, and service account assignments.
Revocation records: When access is revoked, the user who revoked it, the date, and the reason text that user enters.
Asset inventory: Application and device names, hostnames, website domains, criticality classifications, ownership, and asset type metadata.

AI Agents and Non-Human Identities (AI Access)

When you use AI Access, we store records about AI agents, service principals, app registrations, integrations, and other non-human identities in your environment. These records include:

Identity and permission data: Application IDs, display names, granted permissions and consents, app role assignments, and the last time each identity was used.
Ownership data: The employees who own or approved each identity, their acceptance of ownership, and their review decisions with comments. An owner who has no AutoCISO account can act through a secure token link that we send by email.
Evidence packs: Exported records of reviews and decisions. A pack contains the names of reviewers, decision makers, and accountable owners. Only Owners and Managers of your organisation can create and download packs.
Uploaded exports and screenshots: You can upload an export or a screenshot from another system. We read screenshots with text recognition, mask email addresses, and keep only the extracted records. We do not keep the image.

Data from Microsoft Entra ID

If an administrator of your Microsoft tenant grants consent, AutoCISO reads data from Microsoft Graph. This access is read-only. It uses the permissions Application.Read.All and DelegatedPermissionGrant.Read.All, and, only if you enable sign-in activity, AuditLog.Read.All.

We store application and service principal IDs and names, delegated permission grants, app role assignments, and owners. We link owners and consent actors to your employee records.
From sign-in logs, we keep only the application ID and the last sign-in time per resource, for a 30-day window. We discard user names, IP addresses, devices, and locations from those logs.
To end this access, remove the AutoCISO enterprise applications from your Microsoft Entra tenant. Disconnecting in AutoCISO stops collection, but only your administrator can remove the consent in Microsoft Entra.

Compliance & Evidence Data

When you use the ISO 27001 certification and compliance features, we store:

Compliance project data: Scope definitions, gap analysis results, risk register entries, Statement of Applicability records, and readiness scores.
Policy documents: Policy texts, version history, approval decisions, and acknowledgement records.
Evidence artifacts: Files uploaded manually or collected automatically by the Evidence Agent from connected integrations, mapped to specific compliance controls.
Audit records: Internal audit findings, backup test results, and management review records.
Access review data: Campaign configurations, reviewer assignments, and responses (approve/revoke/escalate) submitted by reviewers via secure token links.
Certification data: Uploaded certificates and associated metadata. When you publish a certificate for public verification, the certificate name and issuer are accessible at a public URL.
Privacy programme data: Records of processing activities, data processors, personal data dictionary entries, and data subject request readiness checks that your organisation maintains in the Service.

Supplier Management Data

When you manage suppliers in the Service, we store:

Supplier records: Supplier names, services, contracts, data types handled, sub-processors, and the names, email addresses, and phone numbers of supplier contacts, including privacy contacts.
Assessment data: Questionnaires, answers, attached documents, risk scores, and manual risk adjustments with the justification text your team enters.
External respondents: A supplier contact can answer a questionnaire through a secure token link, without an AutoCISO account. We store the respondent's name, email address, and answers. We use the respondent's IP address to limit abuse of the link.

Risk Management Data

When you use the Risk Management Workspace, we store the records needed to capture, assess, and manage risks across your organisation. This may include:

Risk drafts and records: Risk titles, descriptions, domains, segmentation fields, selected owners, review dates, evidence references, workflow status, and change history.
Assessment and treatment data: Probability, impact, exposure, control strength, estimated loss, treatment cost estimates, selected or recommended treatment strategy, and residual-risk fields.
Risk budget data: Budget amounts, period and scope, consumed and remaining amounts, and plain-language business constraints or assumptions used to derive budget profiles.
AI-guided interview data: Interview sessions, question and answer history, answer options shown, free-text responses, extracted facts, AI interpretations, open questions, summaries, confidence indicators, prompt template version, and model metadata.

AI Chat Data

The AI security advisor sends your prompts to our AI processing pipeline, together with relevant context from your organisation's data. This context can include employee, asset, access, remediation, risk, compliance, privacy programme, supplier, and audit trail records. Conversation history, session identifiers, and folder organisation are stored to enable continuity and sharing of conversations within your organisation.

Integration Credentials

When you connect third-party services, we store the configuration and credentials required to maintain those integrations. Examples are Jira, Microsoft Entra ID, and company identity providers for single sign-on. Credentials are encrypted at rest using AES-256-GCM. Client secrets of company identity providers are passed to our identity platform (Zitadel) and are not kept in the AutoCISO application database.

Audit Trail

All significant actions taken within the Service — including access changes, user administration, policy approvals, compliance events, configuration changes, and reads of personal data through API tokens — are recorded in an audit trail. Audit trail entries include the actor's identity, a timestamp, and a description of the event. Some entries also include the IP address of the request.

Partner Program Data

If you enroll in the vCISO Partner Program, we collect:

Firm data: Your firm or advisor name, firm logo, agreement tier, and the users who manage the partner account.
Portfolio data: Portfolio summary cache records for each managed organization (aggregate counts and security scores only — no individual names, email addresses, or raw access records), and Partner AI Assistant conversation history.
Consent records: For each agreement you accept (program terms, independent contractor terms, anti-corruption representations, and the data processing agreement): the type, version, time, a hashed IP address, and the browser user agent.
Tax and payment data: Legal name, legal address, country, tax identification numbers, VAT status, bank or payout account details, and the tax certificates and invoices you upload. We show only the last digits of tax and bank numbers in the Service and keep the full values encrypted.
Payout records: Referral credit ledger records, statements, VAT amounts, tax withheld, and payment records.
Referral attribution: When someone opens your referral link, we record the referral code, the time, a hashed IP address, the browser user agent, the country, and the referring page.

Browser Extension Data

When you use the AutoCISO Audit Collector browser extension, data is collected only when you explicitly start a capture. Only Owners, Managers, and Contributors of your organisation can submit captures. Specifically:

Simplified page DOM: A stripped-down HTML representation of the SaaS admin page you are viewing, compressed before transmission. Password fields and hidden inputs are automatically excluded before any data leaves your browser.
Screenshots (optional): A scrolled full-page screenshot of the active tab, captured as a compressed PNG, only when you enable this option or when the DOM capture path is unavailable.
Page URL and title: The URL and title of the tab you capture, sent alongside the DOM and screenshot to identify the source application.
Page signals: Lightweight page metadata (page title, meta description, heading text, navigation link text) sent to our API to auto-detect the type of data on the page (users, assets, or access lists). No form values or personal data are included in these signals.

The extension has no standing access to websites. The first time you start a capture on a site, your browser asks you to allow access to that site. The extension runs its capture script on a page only after you start a capture there. You can remove site access at any time in your browser's extension settings.

The extension keeps your session access token in chrome.storage.session, which your browser clears when it closes. Refresh tokens and ID tokens are not stored. The extension keeps no personal data in its local storage, and tokens are never accessible to web pages. In the web dashboard, access tokens are held in memory only and are not persisted to browser storage.

3. How We Use Information

We use your information to:

Provide and maintain the Service, including access inventory, employee lifecycle workflows, compliance management features, supplier management, AI Access, and the risk management workspace.
Serve requests that your organisation's systems and AI clients send to the public API and the MCP server with your API tokens.
Process captured page data through our AI Vision pipeline to detect users, assets, and access relationships.
Power the AI security advisor by providing your prompts and relevant organisational context to our AI processing pipeline.
Generate, store, and update risk drafts, interview sessions, risk records, and budget views so your organisation can identify, assess, prioritise, and treat risks.
Use AI to interpret guided-risk interview answers, extract structured facts, propose draft summaries, and suggest scoring, treatment, and budget-related outputs.
Automate evidence collection and map it to compliance controls within ISO 27001 and other frameworks.
Send supplier questionnaires, access review requests, ownership requests, and other notifications by email.
Track and enforce subscription usage limits (audit runs, AI requests, connected applications, API rate limits).
Communicate with you about your account, security notifications, and product updates.
Ensure security, detect fraud, and maintain the integrity of the Service.
Compute your managed-org discount tier, attribute referrals, and calculate, pay, and report partner payouts, including VAT and tax withholding where the law requires it.
Power the Portfolio AI Assistant, which processes portfolio summary data on your behalf.

Legal Bases for Processing

Where the GDPR or a similar law applies, we rely on these legal bases:

Account, usage and technical data: performance of a contract with you or your organisation, so that we can provide the Service.
Security, abuse prevention, the audit trail and service improvement: our legitimate interests in a secure and working Service.
Sales inquiries: our legitimate interests in replying to a request you sent us.
Website analytics and session recording: your consent, which you give in the cookie banner and can withdraw at any time.
Partner tax and payout records: a legal obligation under tax law.

For workforce, supplier and other customer data that we process as a processor (see section 4), your organisation decides the legal basis as the data controller.

4. Data Controller & Processor Relationship

When you upload or otherwise provide personal data about your employees, contractors, supplier contacts, or other individuals (such as workforce records and access data), your organisation acts as the data controller and AutoCISO acts as a data processor on your behalf. We process that personal data only as directed by you and in accordance with our Data Processing Agreement (DPA), which is available on request. If your organisation is subject to the GDPR or similar data protection legislation, you are responsible for ensuring you have a lawful basis to provide that personal data to AutoCISO.

API and MCP clients. When your organisation connects its own systems or AI clients (for example, an AI assistant connected through the MCP server) with an API token, the Service sends data to those clients at your request. Those clients are under your control. You decide which scopes a token has, and you are responsible for where those clients send or store the data.

Partner Portal Access. vCISO partners access only aggregated, anonymized security posture summaries for managed organizations — not raw tenant data. Individual employee names, email addresses, and access records are never exposed through the Partner Portal. Partners are responsible for their own legal and contractual relationship with their client organizations. AutoCISO is the controller of the partner's own account, consent, tax, and payout data.

5. Data Processing & AI

AutoCISO uses third-party AI models. These models run on the serverless inference service of our cloud provider, DigitalOcean, in its NYC3 region (New York, United States). We send AI models only the data that a feature needs. AI processing is used in these contexts:

Audit pipeline: Simplified HTML and optional screenshots captured via the browser extension are processed by our AI Vision pipeline to extract identity and access information. Text is first read from images by text recognition on our own servers. We mask email addresses before the text reaches an AI model. Other content, such as names, can still be visible to the model. Some screenshots are sent to a vision-capable model when text recognition is not enough. Raw capture data is used solely to produce audit results and is not retained beyond the processing pipeline.
Risk management pipeline: When you use AI-guided or AI-assisted risk features, we send interview prompts, answer content, free-text risk descriptions, and related context needed to generate the next question, extract structured facts, summarize risk scenarios, and suggest scoring or treatment outputs. Those interview records, AI interpretations, extracted facts, summaries, and confidence metadata are stored in your tenant workspace to support resumability, review, and auditability.
AI security advisor (chat): When you use the AI chat feature, your prompt and contextual data from your organisation's records are sent to an AI model to generate a response. Chat history is stored and can be shared within your organisation. AI responses are generated and may not be accurate; they do not constitute professional security, legal, or compliance advice.
Compliance and supplier features: Gap analysis, roadmap generation, evidence collection, policy and record-of-processing suggestions, and questionnaire translation send the relevant records to an AI model.
AI Access: AI analysis of AI agent and non-human identity data runs only if your organisation switches it on. Microsoft Graph data and uploaded exports are processed without AI unless you switch it on.
Portfolio AI Assistant: When you use the Portfolio AI Assistant, we send portfolio-level summary data (counts, scores, trends) to an AI model. No individual names, email addresses, or raw access records are included in AI prompts.

For troubleshooting, AutoCISO system administrators can switch on a short diagnostic capture of AI requests. It is off by default, keeps at most 40 recent requests, stores them encrypted, and switches off automatically.

6. Information Sharing

We do not sell your personal information. We share information only with the following categories of service providers and recipients to operate the Service:

Authentication: Zitadel identity platform (user authentication, multi-factor authentication, and identity management), run by AutoCISO in our own cloud environment.
AI processing: Third-party AI models on the serverless inference service of DigitalOcean in the United States (NYC3 region).
Cloud infrastructure: DigitalOcean (hosting, databases, file and archive storage), NYC3 region, New York, United States.
Email delivery: Brevo, for transactional email such as invitations, access review and ownership requests, supplier questionnaires, reports, and alerts.
Payment processing: The payment provider named on your order or invoice.
Timestamping: Public timestamping servers, which record hashes in the Bitcoin blockchain, receive cryptographic hashes of the audit trail to prove when entries existed. No personal data is sent.
Application logos: Logo.dev receives the website domains of applications in your asset inventory so that the Service can show their logos.
Identity providers (when configured by your organisation): GitHub, Google, or Microsoft, and company identity providers such as Okta, Microsoft Entra ID, or any OpenID Connect or SAML 2.0 provider. These providers tell us whether your email address is verified.
Third-party integrations (when connected by your organisation): Jira receives the title, description, priority, and labels of the risk records, ISO roadmap tasks, and offboarding tasks you send to the Jira site you connect. Microsoft Graph is read from, not written to.
Your own API and AI clients: Systems and AI clients that your organisation connects to the public API or the MCP server with your API tokens.
Tax authorities: Partner payout and withholding data, when the law requires us to report or remit tax.
Sales inquiries: HubSpot, for sales inquiries (the name, work email, company and answers you send through our request forms, stored so we can reply), processes this data in the United States.
Website analytics: Amplitude, for site analytics and session recording on this website, only after you accept cookies. Session recording is off on pages with forms.

Website visitors, request forms and cookies

This applies to people visiting autociso.io, not to users of the Service.

Analytics and session recording: Amplitude loads only after you accept in the cookie banner. If you decline, or never choose, nothing from Amplitude loads. Accepting switches on site analytics and session recording (how you scroll and click). Session recording is off on pages with forms, and on those pages form fields and clicks are not captured. You can change your choice at any time with "Cookie settings" in the footer; declining stops further collection and opts you out.
What the site stores in your browser: cookie-consent and cookie-consent-v in local storage record your choice. If you accept, aca-attribution in session storage keeps the campaign source of your visit (UTM values and whether a partner link was used) until you close the tab, and aca-home-variant in session storage keeps which version of the home page this tab opened (a, b or c) so that a later click on the demo login can be matched to it. Both are deleted when you withdraw your consent. After you accept, Amplitude also sets its own cookies and storage to recognise your browser. Campaign and partner parameters are removed from the address bar when the page loads, and partner codes are never sent to analytics.
Fonts and scripts: The website, the documentation and the web dashboard load fonts and scripts only from our own servers, so opening a page does not send your IP address to a font or script provider.
Partner referral links: When you open a partner's referral link, we set the first-party cookie auc_ref so that a later sign-up can be credited to that partner. We record the click as described under Partner Program Data. This cookie is needed to pay partners and is not used for analytics.
Request forms: When you send a request form, we send your answers to HubSpot, and we use them only to reply and to plan the review you asked about. The forms do not set HubSpot tracking cookies. Free-text notes should not contain passwords, keys or other people's personal data.
Retention of inquiries: We keep inquiry records for 12 months after last contact for people who do not become customers, then delete them. We delete an inquiry earlier on request: write to legal@autociso.io.

We may also disclose information where required by law or to protect the rights, property, or safety of AutoCISO, our users, or others.

International Transfers

Some of our service providers process personal data outside the European Economic Area, mainly in the United States. For these transfers we rely on the providers' data processing agreements, which include the EU Standard Contractual Clauses, or on the EU-U.S. Data Privacy Framework where the provider is certified. You can ask us for a copy of the relevant safeguards at legal@autociso.io.

7. Public-Facing Features

The Service includes features that make certain data accessible outside your organisation's account. When you publish an ISO certificate for public verification, the certificate name, issuer, and validity dates become accessible at a public URL. When you share an auditor portal link, an access review link, a supplier questionnaire link, or an AI agent ownership or review link, the recipient can access the relevant data without logging in, for as long as the token remains valid. You are responsible for managing access to these links and revoking them when no longer needed.

8. Data Retention

We keep your organisation's data while your subscription is active, unless a shorter period applies below:

Audit trail: 7 years by default. Entries older than 90 days move to encrypted archive storage.
Deleted organisations: When an organisation is deleted, its data is kept for 30 days so that the deletion can be reversed, and is then permanently deleted.
AI Access collection runs: The latest 3 complete runs per source. Older runs are removed after 30 days.
Evidence packs: 90 days by default. Your organisation can set a period from 7 to 365 days.
Product usage events: 400 days.
AI chat history: Until you delete the conversation or the organisation is deleted.
Partner tax and payout records: 7 years, or longer if tax law requires it. We cannot delete these records earlier on request.
Referral click records: 18 months.
Sales inquiries: 12 months after last contact, as described above.

9. Your Rights

Depending on your location, you may have rights under the GDPR, CCPA, or other privacy laws to access, correct, or delete your personal data, to restrict or object to its processing, to data portability, and to withdraw your consent at any time without affecting processing done before the withdrawal. If you are an employee, supplier contact, or AI agent owner whose data has been entered by an organisation using AutoCISO, please direct requests to that organisation as the data controller. AutoCISO users and partners may contact us directly at legal@autociso.io to exercise their rights. You also have the right to lodge a complaint with a data protection supervisory authority, for example in the country where you live or work.

Children

The Service is for businesses and is not directed to children, and we do not knowingly collect personal data from anyone under 16. If you believe a child has given us personal data, write to legal@autociso.io and we will delete it.

10. Security

We use AES-256-GCM encryption for stored credentials and sensitive data at rest, TLS for data in transit, and continuously scan dependencies against the NVD vulnerability database. Our identity platform supports multi-factor authentication. Access tokens in the web dashboard are held in memory only and not written to browser storage. API tokens have limited scopes and an expiry date. Only an Owner can create a token that reads personal data, and such a token expires after at most 90 days. Role-based access control and multi-tenant isolation ensure that each organisation's data is accessible only to authorised users within that organisation.

11. Contact Us

If you have any questions about this Privacy Policy or wish to request a Data Processing Agreement, please contact us at legal@autociso.io.