Your information
Privacy policy
How Spark Personal Assistant accesses, uses, stores, and shares information.
Effective and last updated: October 11, 2026
Who operates Spark
Spark Personal Assistant is a private personal assistant operated by Jason Kahn through Kahn Creations LLC, 10 Crawfords Corner Road, STE 219, Holmdel, NJ 07733, United States. This policy covers Spark’s private workspace and its connected Google accounts. Other tools on JasonKahn.com have a separate privacy policy.
For questions, access requests, or deletion requests, contact hello@jasonkahn.com.
Information Spark accesses
Spark accesses only accounts you authorize and the features enabled for those accounts. Google’s consent screen shows the permissions requested.
- Gmail: your connected email address; message and draft content; senders, recipients, subjects, dates, labels, and conversation identifiers; and attachment information. Attachments may also be processed when included in a managed draft or an authorized task. These permissions support reading and summarizing mail, preparing and editing drafts, organizing messages, and sending an exact message after your approval.
- Google Calendar: connected calendar identifiers and event details, including titles, descriptions, times, locations, attendees, response status, and meeting links. These support schedule views, reviewed event changes and invitation responses, and configured synchronization of busy blocks across your accounts.
- Google authorization: OAuth tokens and the granted permissions needed to keep the selected accounts connected. Spark does not receive your Google password.
- Workspace information: your messages to Spark, uploaded or selected files, preferences, saved memories, tasks, project context, approvals, action results, and activity records. These can contain excerpts or summaries of connected Google data.
- Access and device information: sign-in and passkey records, session cookies, browser storage, notification subscriptions if enabled, and operational or security records needed to provide remote access and diagnose failures.
How the information is used
Information is used to provide the assistant features you request: answer questions, find relevant past work, summarize mail, prepare drafts, organize messages, show your schedule, manage tasks and projects, carry out reviewed actions, and run routines you enable. Background synchronization, draft preparation, mailbox organization, and calendar mirroring may continue while the app is closed when you have enabled those features.
Google permissions can include broader capabilities than Spark makes available. Spark uses application controls to limit automatic work. Sending email and reviewed calendar actions use separate approval flows; existing synchronization and housekeeping follow your configured settings. You can revoke Google access at any time.
Google data and information derived from it are not sold, used for advertising, used to determine creditworthiness, or used to train general-purpose AI models.
Local processing and optional AI assistance
Routine model processing runs on the operator’s private computer. Relevant message or calendar excerpts may be included in a model request to answer your question or perform an enabled assistant feature. Saved memory supports retrieval across conversations; it is not model training.
If you choose optional cloud assistance, Spark prepares a review of the material to share. Only after a separate approval is that material sent to the configured external provider, currently Anthropic’s Claude. The review may include your request, selected context, and prior action results. Do not approve a packet you do not want that provider to process. Google data shared for this feature remains subject to the Limited Use commitments below.
Storage, service providers, and sharing
Google tokens, synchronized content, conversations, memories, project records, and action history are stored on systems controlled by the operator, primarily the private assistant computer. Cloudflare hosts the website and remote access service, handles authenticated requests and responses in transit, and stores the session and operational information needed for that service. This is not a claim that all information stays on your phone or that the hosting provider can never process it.
Google processes authorized API requests. If you enable notifications, your browser or device push service delivers them; previews may reveal assistant content on your lock screen. Connected messaging services process messages delivered through a channel you choose. Optional external AI processing is described above. A message or event action you approve shares the reviewed information with its intended recipients.
Information may also be accessed or disclosed when needed to protect security, investigate abuse, or comply with applicable law. Human access to Google data is limited to your affirmative permission to view specific information, necessary security or legal purposes, or other uses expressly permitted by Google’s Limited Use requirements. It is not shared with advertisers or data brokers.
Retention, revocation, and deletion
Connected content and saved workspace records are retained while needed for the features you use or until you request deletion. Conversations, memories, drafts, and action history can persist across restarts. There is no automatic promise that closing the app, removing a message in Google, or disconnecting an account erases every previously saved copy or summary.
To stop future Google access, remove Spark’s connection on your Google Account connections page. Revoking access stops authorized API access once Google invalidates the grant; it does not by itself delete existing local copies.
To request a copy or deletion of data stored by Spark, email hello@jasonkahn.com and identify the connected account and the information concerned. We may verify control of the account before fulfilling the request. Requests cover applicable synchronized copies, saved memories and conversations, and retained credentials. Limited records may need to be retained for security, resolving an unfinished action, or legal obligations. Backup copies are included in the deletion review and may remain until the relevant backup is removed or replaced.
Deleting Spark’s local copy does not automatically delete the original email or event in your Google account. Changes to original Google data require the relevant authorized action.
Security and your choices
Spark uses authenticated private access, passkeys where configured, restricted access to local data and credential files, and HTTPS for external service connections. No system can guarantee complete security. Keep your devices and Google account secure, review proposed actions, and turn off notification previews if you do not want them displayed on a locked device.
You may decline a Google connection, revoke it, decline a cloud handoff, disable device notifications, or request deletion. Some features will be unavailable without the permissions they need.
Google API Limited Use
Spark Personal Assistant’s use and transfer of information received from Google APIs adheres to the Google API Services User Data Policy, including the Limited Use requirements. This applies to the original Google data and information derived from it.
Access and transfers are limited to providing the visible assistant features with your consent, necessary security purposes, applicable legal obligations, and other cases expressly permitted by that policy. Google user data is not used for advertising or to train general-purpose AI models.
Children and policy updates
Spark is a private tool for authorized adults and is not directed to children. If you believe a child’s information has been provided, contact us to review and remove it.
Updates will be posted here with a new date. Material new uses of Google data will be disclosed and require any additional consent before they begin.