How we use AI
How automated summaries are produced, checked and labelled.
Last updated
What AI assists with
Google Gemini helps select financially relevant news, group related headlines and draft summaries and context. Automated validation checks fields, categories, locations and output limits before a saved feed is published. This is automated processing with validation, not a claim that a human approves every article.
What the model sees
The news workflow supplies public source headlines and short descriptions and reference information used to organise the news. It does not send account emails, private watchlists or visitor messages to Gemini. Summaries link to their original sources rather than reproducing full articles.
Limits and transparency
Models can misread context, omit important details or produce errors. Validation cannot establish that every statement is true. AI transparency here follows the spirit of informing readers, including the objectives of EU AI Act article 50; this is not a claim of certified compliance or a substitute for checking which legal obligations apply.
Images and corrections
Topic photographs from Pixabay are labelled “Illustration”; they may not show the event described. They are not presented as eyewitness evidence or automatically assumed to be AI-generated. Logos identify their owners. Report suspected errors to [email protected] [OWNER: confirm mailbox forwarding and monitoring] with the article URL, disputed passage and a reliable source. See the corrections process.