AI and your data

Random uses AI models to sort mail, write drafts, briefs and prep, and answer in Chat. The requests go to EU data centres, and your data is not used to train AI models.

Where AI requests go

Requests go through an AI routing service on its EU endpoint, to model providers in EU data regions. The routing service keeps request metadata, not content. The company behind it is based in the US. The model providers are listed in Annex III of our data processing agreement.

Dictation is transcribed by a provider in the EU. While you dictate, the live words come from an AI provider in France (EU); some of its own sub-processors may process data outside the EEA, under standard contractual clauses. It keeps audio and text for up to 30 days for abuse monitoring and never uses them for training. See the sub-processor register.

No training on your data

Your data is not used to train AI models. Random learns your style and your contacts inside your own workspace, and you can see, correct or forget what it learned in Random Brain.

Web research

Web research is off by default. When you turn it on, Chat can search the web for you. The search query and a market code go to a search provider in France that may fall back to a provider in the US; pages it reads are fetched from their sites.

Fair use and your review

The plan covers AI use under fair use, with no credits to count. If a workspace reaches the ceiling, AI work pauses until it resets. See Plan, trial and fair use.

Drafts, briefs and summaries are written by AI models. Review them before you use them.