SAL AI Privacy Policy
**Last updated:** October 09, 2025
**Applies to:** SAL AI CustomGPT (“Service”) and the following public Actions: Google Scholar (via SerpAPI), OpenAlex, arXiv Search, Readable Web Fetch (Jina Reader), GitHub Repository Search, Semantic Scholar.
---
1) Who we are
**Controller/Operator:** SAL AI (operated by Sal Beas).
**Contact:** info@sal-ai.com
This Privacy Policy explains what data the Service processes when you use SAL AI CustomGPT, how it’s used, the third-party APIs we call, and your choices.
---
2) What this covers
This policy applies to the SAL AI CustomGPT hosted on OpenAI’s platform and its Actions (APIs) listed above. It does **not** cover:
- OpenAI’s processing (see OpenAI’s own privacy policy);
- Third-party sites you visit by clicking links returned in results;
- Any services you use to host this policy (e.g., GitHub Pages, Notion).
---
3) Data we process
### A. You provide
- Prompts, questions, and file contents you upload (e.g., PDFs).
- Action inputs (e.g., search queries, filters, parameters).
### B. Automatically from your use
- Timestamps, basic usage metadata (e.g., which Action was invoked).
- Technical request logs generated by OpenAI and third-party APIs (see §7).
We **do not** deliberately collect sensitive data. Please avoid including health, financial, precise location, or other sensitive information in prompts unless strictly necessary.
---
4) What each Action sends/receives
The Service can call external APIs to fetch public information. When you ask it to search or fetch content, the following may be sent to the corresponding provider and responses are returned to you in the chat.
- **Google Scholar (via SerpAPI)** — Sends your query and optional filters to SerpAPI; receives structured Scholar results (titles, links, snippets, metadata).
- **OpenAlex** — Sends your search terms and filters; receives scholarly metadata (works, institutions, DOIs, authorship, venues).
- **arXiv Search (Atom)** — Sends your query and sort parameters; receives Atom XML entries (title, authors, summary, links).
- **Readable Web Fetch (Jina Reader)** — Sends a public URL; receives readable/plaintext extraction of that page.
- **GitHub Repository Search** — Sends your search terms; receives repository metadata (name, URL, description, stars).
- **Semantic Scholar** — Sends your search terms and requested fields; receives paper metadata (title, authors, venue, abstract, URL).
**Note:** These providers may log your request (including IP address as seen from OpenAI’s infrastructure), standard HTTP headers, and timestamps per their policies.
---
5) How we use data
- To fulfill your requests (e.g., run searches, summarize results, cite sources).
- To improve response quality during your session (contextual understanding).
- To prevent abuse (e.g., detect harmful prompts using safety heuristics).
We do **not** sell your data. We do **not** profile you for advertising.
---
6) Legal bases (if you are in the EEA/UK/Switzerland)
- **Contract**: Processing needed to provide the Service you request.
- **Legitimate interests**: Abuse prevention, ensuring availability, measuring basic usage.
- **Consent**: If you explicitly ask us to process special categories of data.
---
7) Sharing with third parties
We share data only as needed to fulfill your request with these third-party APIs:
- SerpAPI (Google Scholar): https://serpapi.com
- OpenAlex: https://api.openalex.org
- arXiv: https://export.arxiv.org
- Jina Reader: https://r.jina.ai
- GitHub API: https://api.github.com
- Semantic Scholar: https://api.semanticscholar.org
Each provider’s own privacy policy applies to their handling of requests and logs.
We also rely on **OpenAI** to operate the chat interface and model inference. See OpenAI’s privacy documentation for model and log handling.
---
8) Data retention
- **Within the CustomGPT**: We don’t create separate databases. Conversation content may persist within OpenAI’s system as governed by OpenAI.
- **Third-party APIs**: Providers may retain request logs per their policies.
- **Your uploads**: If you upload files to the chat, they remain accessible in the session/account until you delete them within OpenAI’s interface.
If you need content removed from a GitHub Pages/Notion site where this policy is hosted, contact us at **info@sal-ai.com**.
---
9) Security
We use reputable providers (OpenAI and the listed APIs) that support HTTPS/TLS. No method of transmission or storage is 100% secure; submit only what you’re comfortable sharing.
---
10) Your rights
Depending on your location, you may have rights to access, correct, delete, or restrict processing of your personal data. Contact **info@sal-ai.com** to exercise your rights. For data held by OpenAI or third-party APIs, please contact them directly.
---
11) International transfers
APIs may be hosted globally. Your requests may be processed outside your country. We rely on providers’ transfer mechanisms and safeguards.
---
12) Children
The Service is intended for users **13+** (or the age required by local law). Do not use the Service if you are under that age.
---
13) Changes
We may update this policy as features or laws change. Material changes will be reflected by updating the “Last updated” date above.
---
14) Contact
**SAL AI (Sal Beas)**
Email: **info@sal-ai.com**