AI Act Art. 52 transparency disclosures: how do you prove compliance during an audit?

In our organization we deployed several AI-powered features: a customer-support summarizer, an internal document classifier, and an employee-facing scheduling bot. Under the EU AI Act, all three arguably fall under Art. 52 transparency obligations (users must be informed they're interacting with AI). The tricky part isn't the disclosure itself — it's the audit trail. How are teams documenting: 1. That the disclosure was actually presented to the user (not buried in ToS)? 2. The timing and context of each disclosure (pre-interaction vs. inline)? 3. Handling multilingual/multi-jurisdictional deployments where the AI Act interacts with national laws? We're debating between: (a) a centralized consent/disclosure log, (b) embedding disclosure timestamps in each product's own event stream, or (c) relying on UI-level screenshots during compliance audits. None feel fully robust for an auditor. What's worked in practice for proving Art. 52 compliance during an audit? Jurisdiction: EU, DE, AGNOSTIC