AI Transparency
Clear, honest, and built for court scrutiny. Here's exactly where the AI's information comes from — and just as importantly, where it does not.
The short answer
The AI only knows what you put in the case. It has no access to the internet, no outside databases, and no way to look anything up on its own.
The AI reads only the material you add to a case — your evidence items, notes, clue descriptions, and attached documents. That's the entire universe of information it has access to. Nothing more.
Clueva's AI does not search the web, browse websites, or pull from Google. It cannot look anything up on its own. If it's not in your case, the AI doesn't know about it.
No criminal records, no DMV data, no background-check feeds, no people-search services, no government databases. Clueva is not connected to any of them.
Clueva has no tools to scrape social media, harvest profiles, or collect data from any platform. If you want to reference a public post, you upload it yourself as evidence.
Clueva does not run facial recognition, biometric matching, or identity lookups on any image. It describes what's visibly in a photo you provide — it never tries to identify who someone is.
An investigator adds material — a witness statement, a report, a document, a photo, or notes. This is the only information the AI will ever work from.
When you click a feature like "Extract Clues" or ask the agent a question, the relevant case content is sent over an encrypted connection to our AI provider (Anthropic's Claude) to be analyzed.
Claude reads the text or image you provided and returns structured results — extracted clues, suggested connections, answers about your documents, or a draft summary. It reasons only over the content you sent.
Nothing the AI produces is final. Every extracted clue, every suggested connection, and every narrative is a suggestion that an investigator reviews, edits, or rejects before it becomes part of the case.
Clueva is built on Anthropic's Claude — a leading large language model. Claude reads the text and images you provide and produces structured analysis. It is a reasoning model, not a database, and it has no independent access to outside information.
No. Your case content is sent to the AI provider only to process your request and return a result. It is not used to train or improve any AI model. Your investigative materials stay yours.
No. The AI has no memory of other cases and no connection to outside data. If you ask about a person, it can only reflect what your own uploaded evidence says about them — described by observable attributes, never identified as a suspect or scored.
Every detail traces back to a piece of evidence you uploaded. Clueva is built so that any claim in a report can be followed to its source. The AI does not add facts that aren't in your case.
Yes — AI output can be incomplete or inaccurate, which is exactly why a human reviews everything. Clueva is a tool to help investigators work faster, not a source of truth. Always verify against the original evidence.
Try the live demo — upload a document and ask the agent about it. You'll see it only ever answers from what's in the case.