AI Disclosure

Effective date: April 18, 2026

Probed (probed.space) is an intelligence platform that aggregates, clusters, and analyzes publicly available reporting on UAP (Unidentified Anomalous Phenomena) and adjacent topics. AI assists in nearly every stage of how that information is processed and presented. This page explains, in plain language, where AI is used, what is and isn't generated automatically, how we mitigate errors, and what your rights are when AI-generated content concerns you.

1. Where AI is used

We use commercial large language models, vector embeddings, and conventional algorithms from one or more third-party AI providers in the following parts of the Service. The specific models in use evolve over time as the underlying systems improve; this disclosure covers how AI is applied rather than which model is invoked at any moment.

Item classification. Each ingested item (article, video, post, document, sighting report) is automatically categorized for relevance, topic, evidence type, and credibility signals.
Entity extraction and linking. Mentions of people, organizations, events, locations, and topics are extracted and linked to first-class entity records.
Signal clustering. Related items are grouped into "signals" using vector similarity and topic features. Each signal receives a score derived from source corroboration, freshness, and evidence type.
Signal titles. When a signal first forms, an AI-generated headline is assigned to summarize the cluster. As more sources arrive, the title may be revised automatically to better reflect the corroborated story.
Summaries and analysis. Signal pages, entity pages, and sighting pages may include AI-assisted summaries, narrative analysis, key-claim extraction, timelines, and related-entity sections. These are clearly labeled within the page.
Deep ingestion of long-form content. When a long article, post, or transcript is processed end-to-end, the resulting structured breakdown (claims, evidence, timeline events, related entities) is AI-generated and shown alongside a prominent disclaimer on the item page.
Daily digests and briefs. Periodic email and in-product digests are typically system-generated summaries of recent signals.
Articles and editorial content. Some articles begin as AI-generated drafts and are subsequently reviewed and edited by humans before publication. See our Editorial Standards page for more on the human-in-the-loop process.

2. Where AI is not used

Quotes come from underlying sources. Quoted text on the platform is drawn from the source material we ingest and is attributed to that source. AI may select, excerpt, or position those quotes within a summary, but our pipelines are not designed to author quotations that do not appear in the source. If you spot a quote that appears fabricated or materially altered, please report it (see Reporting an error).
We do not use AI to make moderation, account, or enforcement decisions about you in a way that produces legal effects without human review.
We do not use your personal information to train or fine-tune AI or machine-learning models. See our Privacy Policy for details.
Sighting reports submitted by users are stored and displayed as submitted, with environmental context (weather, geomagnetic, satellite passes, astronomy) layered on top algorithmically. The witness account itself is not rewritten.

3. How AI-generated content is labeled

Signal titles carry an information icon (ⓘ) next to the title indicating the title is AI-generated and may be revised as new reporting arrives.
Probed Analysis sections on signal, entity, and sighting pages carry an information icon (ⓘ) indicating the analysis is AI-assisted and that readers should consult the original sources for full context.
Deep-ingest pages for long-form items show a prominent disclaimer card at the top of the structured breakdown.
The information icons above link directly to this page, and this page is also reachable from the About page in the navigation footer on every screen of the Service.

4. Accuracy, errors, and limits

AI systems make mistakes. Despite our scoring, clustering, and attribution machinery, you may encounter:

Titles that don't fully capture the most important framing of a story.
Summaries that compress nuance, omit context, or carry inferences that the original source didn't make.
Entity linking that is incorrect (e.g., name collisions, confused organizations).
Cluster boundaries that merge or split stories in ways a human editor might do differently.

We continuously tune the underlying systems and surface confidence and source-count signals on the page so readers can calibrate trust. We encourage you to follow the source links on every signal and item — the structured layer is meant to help you find primary reporting faster, not replace it.

5. Reporting an error or requesting correction

If an AI-generated headline, summary, classification, or entity link concerns you — particularly if you are the subject of the characterization or believe it is materially inaccurate — email probedspace@gmail.com with the URL of the page in question and a brief description of the issue. We aim to review correction requests within a reasonable timeframe and will revise, remove, or annotate the affected content where appropriate.

6. Compliance posture

This disclosure is intended to support, among other things, transparency obligations under the EU AI Act (Regulation (EU) 2024/1689), the U.S. Federal Trade Commission's guidance on truthful representation of AI-derived content, and analogous emerging regimes. AI-generated and AI-assisted content on the Service is labeled as such on the artifact itself and through this page.

The Service uses generative AI to produce text and structured metadata; it does not generate synthetic media designed to imitate a real person's likeness or voice (deepfakes), and it does not use AI to make decisions that produce legal effects about you.

7. Changes to this disclosure

We may update this AI Disclosure as the Service's use of AI evolves and as the underlying regulatory landscape matures. When we make material changes we will revise the effective date at the top of this page and, where appropriate, surface a notice within the Service.

8. Contact

Questions about how we use AI or about a specific piece of AI-generated content can be sent to probedspace@gmail.com.