Last Updated: March 23, 2026
TL;DR: When you click the Gist button, we process the article or video through paid AI services, cache the summary, and delete the original. If you don't interact with Auto-Suggest, nothing leaves your device. If you interact with it (click or dismiss), we collect page metadata including the URL — but never the page content — to improve recommendations for you and all users. If — and only if — you enable Debug Mode, you are in the Eye of Sauron. You can disable Auto-Suggest entirely and blacklist specific domains. Our AI providers are on paid tiers and are contractually prohibited from training on your data.
Who We Are
Mireles & Co, LLC, dba Transcendence, Inc. ("we," "us," "our") operates the Beta version of Gist, a browser extension that helps you read faster by summarizing articles and videos.
Contact us at: privacy@gist.is
How Gist Handles Your Data
Gist has two core features — Gist summaries (which generate AI-powered summaries of articles and videos) and Auto-Suggest (which recommends content to summarize) — plus a diagnostic tool called Debug Mode. Each has a different data profile.
Gist Summaries
When you click the Gist button to summarize an article or video:
1. The full source content is sent to our servers
2. We route it to our AI providers (Google Gemini and Anthropic Claude via AWS Bedrock) for summarization
3. The AI-generated summary is returned and cached in your content library
4. The original source content is deleted from our servers
We keep the summary. We do not keep the content you scraped. We do not use source content to train AI models during standard use.
Auto-Suggest
Auto-Suggest recommends content worth summarizing. It shows a bar at the bottom of the page when it thinks you'd benefit from a Gist.
If you ignore it: All processing stays on your device. Zero data transmission. We don't know what page you're on, what you're reading, or that you're even using Gist.
If you interact with it — by clicking it, dismissing it, or adjusting its settings — we collect:
- Page metadata: The URL, DOM structure features (such as number of links and layout characteristics), and similar structural signals. We do not collect the actual text content of the page.
- Interaction signal: Whether you clicked, dismissed, or otherwise engaged with the suggestion.
How we use this data:
- Personalization: We use your interaction history to improve which pages Gist suggests summarizing for you.
- Collaborative filtering: We use interaction patterns in aggregate across all users to improve recommendation quality globally. This means your interaction data (in anonymized, aggregate form) helps shape what other users see, and theirs helps shape what you see.
- Model training: We train a recommendation model using this interaction data and page metadata to improve Auto-Suggest accuracy.
What we do NOT collect at this stage: The actual content of the page. We see structural features of the page, not what the page says.
Domain Blacklisting
If you blacklist a domain from Auto-Suggest through your settings, we collect the blacklisted domain. We use this to suppress recommendations on that domain for you and, in aggregate, to improve recommendations globally.
A note on sensitive domains: Blacklisting a domain tells us you visit that domain. If a domain is sensitive to you (medical, financial, personal, or otherwise), be aware that we will have a record of the blacklisted domain associated with your account. If this concerns you, you can disable Auto-Suggest entirely instead of blacklisting specific domains.
Debug Mode (Opt-In Only)
Debug Mode is a diagnostic feature that is never enabled by default. If — and only if — you enable Debug Mode, you are in the Eye of Sauron. We may ask you to enable it when you contact support. See the dedicated Debug Mode section below for full details.
Your Controls
You have fine-grained control over how Gist collects data:
- Disable Auto-Suggest entirely: No Auto-Suggest interaction data is collected
- Blacklist specific domains: Auto-Suggest won't appear on those domains (the blacklisted domain itself is collected)
- Tune Auto-Suggest behavior: Adjust sensitivity and preferences in settings
- Disable Debug Mode: Stop enhanced data collection at any time
- Uninstall the extension: Stop all data collection immediately
What We Always Collect
Regardless of which tier you're in, we collect:
Account Information
- Email address (via Google Login)
- Firebase authentication tokens
Product Analytics (Amplitude and Cloudflare Analytics)
- Feature usage (which Gist features you use and how often)
- Session data (how often you open the extension, general usage patterns)
- Performance metrics (load times, processing speeds)
- General device and browser information
- Error rates and crash data
These analytics tell us which features work and which don't. They track your usage of Gist features. They do not track the content of pages you visit or your general browsing behavior.
Debug Mode
Debug Mode is a diagnostic feature for troubleshooting issues you report to our support team. If — and only if — you enable Debug Mode, you are in the Eye of Sauron.
How It Works:
- You enable Debug Mode manually through a clear, explicit opt-in process
- A prominent banner is displayed on your settings page while Debug Mode is active
- You can turn it off at any time
What changes when Debug Mode is on:
- We retain the full source content you scrape (stored in Cloudflare R2) instead of deleting it after processing
- We track the pages you visit and detailed session activity
- We may use data collected during Debug Mode to train, fine-tune, or evaluate AI models and on-device heuristics
What happens when you turn Debug Mode off:
- Additional data collection stops immediately
- Data already collected during Debug Mode is retained for debugging and model improvement
Why it exists: Debug Mode captures the data we need when our on-device heuristics or AI models malfunction. If something is broken, we need real examples to diagnose and fix it.
Third-Party Services
| Service | Purpose | What They Receive |
|---|---|---|
| Firebase / Google Cloud | Authentication, database | Account info, cached summaries |
| Google Gemini (paid tier) | AI summarization | Source content (deleted after processing). Contractually prohibited from training on your data. |
| Anthropic Claude via AWS Bedrock (paid tier) | AI summarization | Source content (deleted after processing). Contractually prohibited from training on your data. |
| Cloudflare | Hosting, edge computing, CDN, analytics, storage (R2) | Cached data, aggregated analytics, Debug Mode data (when applicable) |
| Amplitude | Product analytics | Feature usage, session data |
| Stripe | Payment processing (when applicable) | Payment information |
Important: We use paid API tiers for both Google Gemini and Anthropic Claude. Under these agreements, our AI providers are contractually prohibited from using your data to train their models.
How We Use Your Data — Summary Table
| Data Type | Used For | Shared With Others? |
|---|---|---|
| Auto-Suggest interaction data | Personalization, collaborative filtering, model training | Yes — in anonymized, aggregate form to improve recommendations for all users |
| Blacklisted domains | Suppressing recommendations for you; in aggregate, improving recommendations globally | Yes — in aggregate only |
| Source content (Gist click) | Generating summaries | AI providers only (for processing); deleted after summary generation |
| Cached summaries | Your content library | Not shared |
| Product analytics | Feature improvement, bug detection | Analytics providers (Amplitude, Cloudflare) |
| Debug Mode data | Diagnosing issues, model training/evaluation | Not shared externally |
Your Rights
What You Can Do Now
- Disable Auto-Suggest to prevent Auto-Suggest interaction and blacklist data collection
- Uninstall the extension at any time
- Disable Debug Mode at any time
- Request a copy of your data by emailing privacy@gist.is (within 30 days)
- Request deletion of your cached summaries by emailing privacy@gist.is
- Ask questions about our data practices at any time
Post-Beta
When we exit beta, we'll implement standard data controls including self-service deletion, data portability, and granular privacy settings.
Data Retention
| Data Type | Retention Period |
|---|---|
| Account information | Duration of your account |
| Cached summaries | Duration of your account (deletion available on request) |
| Auto-Suggest interaction data | Used for model training; retained in aggregate |
| Blacklisted domains | Duration of your account |
| Product analytics (Amplitude/Cloudflare) | Per provider defaults (typically 12–24 months) |
| Source content (standard use) | Deleted immediately after summary generation |
| Debug Mode data | Retained for debugging and model improvement |
Security
We implement reasonable security measures including:
- Encrypted data transmission (HTTPS everywhere)
- Encrypted storage (Firestore, Cloudflare R2)
- Authentication required for all data access
- Paid AI provider tiers with contractual data protections
- Regular security updates
We're a small beta product. We take security seriously, but we're not making promises we can't keep.
Age Restrictions
You must be at least 16 years old to use the Beta version of Gist. We do not knowingly collect data from anyone under 16. If we learn we have, we'll delete it.
International Users
We're a California LLC but accept users globally. By using Gist, your data will be processed in the United States.
GDPR (European Users)
Our legal basis for processing varies by data type:
- Auto-Suggest interaction data and collaborative filtering: Legitimate interest (improving service quality). You can opt out by disabling Auto-Suggest.
- Content processing (Gist click): Consent, provided by clicking the Gist button.
- Debug Mode: Explicit consent, provided through the opt-in flow.
You have the right to complain to your local data protection authority.
CCPA (California Users)
We do not sell your personal information. You can request disclosure of the categories of information we collect by emailing privacy@gist.is.
Cookies and Tracking
- The extension itself does not use cookies
- Our website (gist.is) uses cookies for authentication
- We use local storage in your browser for client-side caching and on-device processing
- Amplitude and Cloudflare Analytics operate server-side
AI-Specific Disclosures
Auto-Suggest Model:
- We train a recommendation model using page metadata and interaction signals
- This model runs on-device for suggestions, but is trained using data collected from user interactions
- Interaction data contributes to collaborative filtering that improves recommendations for all users
Summarization (Standard Use):
- Your source content is sent to Google Gemini and Anthropic Claude for summarization
- Both providers are on paid API tiers that contractually prohibit them from training on your data
- Source content is deleted from our servers after summary generation
- We do not use source content to train AI models during standard use
Debug Mode:
- Source content may be retained and used to train, fine-tune, or evaluate our AI models and on-device heuristics
- This applies only to data collected while Debug Mode is explicitly enabled
Data Breaches
If there's a breach that affects your data, we'll notify you within 72 hours of discovery via email and in-app notification.
Changes to This Policy
We'll update this policy as needed. We'll notify you of material changes via email, in-app notification, or update notice in the extension. During the beta period, we may make minor changes without notice.
The Bottom Line
We built Gist to minimize data collection by default. When you click the Gist button, we process the content, keep the summary, and delete the original. If you don't interact with Auto-Suggest, nothing leaves your device. When you interact with Auto-Suggest, we collect page metadata to improve recommendations — but never the page content itself. If — and only if — you opt into Debug Mode, you are in the Eye of Sauron. You have controls over all of this. Our AI providers are paid and contractually barred from training on your data.
When we exit beta, you'll get even more control. For now, this is how it works.
Contact Us
Email: privacy@gist.is
Mailing Address:
Mireles & Co, LLC dba Transcendence, Inc.
506 Island Ave
Reno, NV 89501
---
This policy was last updated on March 23, 2026 and applies to the Beta version of Gist.