10 Audit Evidence Examples and How to Document Them
Ten audit evidence examples explained, from source documents to AI verification trails, with guidance on documenting source, proce…
31/10/2026
Governance
10 Audit Evidence Examples and How to Document Them
Ten audit evidence examples explained, from source documents to AI verification trails, with guidance on documenting source, procedure, result and limitation.
Latest
All articles →- 30/10/2026 University of Helsinki finds measurement errors
- 29/10/2026 METR and Redwood found blind spots in OpenAI
- 28/10/2026 Model Router spreads requests across AI providers
- 28/10/2026 AI Act machine-readable marking but watermarks fall short
- 27/10/2026 EACL study: prompt language and model origin
- 27/10/2026 Red teaming of AI agents must attack the monitor
- 27/10/2026 NIST, NSA and CISA practices for AI incident response
- 26/10/2026 Bank of England and FCA report growing third-party AI
- 26/10/2026 An audit right in AI contracts: specify the evidence
- 25/10/2026 AI explanation is not proof: record the decision
Governance
All Governance →
University of Helsinki finds measurement errors
Research from the University of Helsinki and ACL studies shows hallucination detection must measure factuality, source faithfulnes…
30/10/2026
METR and Redwood found blind spots in OpenAI
Red teaming of agent monitoring and academic research show that external tool calls expand an AI agent's powers; the EU AI Act als…
29/10/2026
Risk
All Risk →
Red teaming of AI agents must attack the monitor
METR, Anthropic and OWASP show that prompt injection in AI agents can only be tested defensibly through a structured assessment of…
27/10/2026
AI voices in Rotterdam bank helpdesk fraud
Rotterdam police arrested two suspects for bank helpdesk fraud using AI voices and NFC abuse. What this demands of decision-makers…
22/10/2026
How many repeats a prompt reliability test needs
New peer-reviewed research into LLM classification shows that reproducibility tests must be designed statistically up front: effec…
14/10/2026
Data
All Data →
When is an anonymised dataset truly anonymous with AI?
The EDPB's 2026 guidelines test anonymity against record isolation, linkability and inference. Read why AI forces this framework a…
23/10/2026
Voiceprint lawsuit makes voice-data provenance governance
Seven Illinois voice professionals sue Microsoft over voiceprints from public recordings. What this means for provenance, consent,…
20/10/2026
Study shows LLMs leak personal data in notes
Research shows LLMs can leak third-party privacy-sensitive data despite privacy instructions. How to separate retrieval, sanitisat…
19/10/2026
Policy
All Policy →
Marking AI content from 2 August 2026: what the AP expects from your workflow
From 2 August 2026 the Dutch Data Protection Authority supervises AI transparency. What must you label, watermark and demonstrably…
04/09/2026
EU AI Act: fines up to 15 million euros for providers of GPAI models from 2 August 2026
From 2 August 2026 the AI Office can fine providers of GPAI models such as ChatGPT, Claude and Gemini up to 15 million euros or 3%…
02/09/2026
Anonymised data is no longer an end state after EDPB Guidelines 02/2026
The EDPB Guidelines 02/2026 make re-identifiability after anonymisation a testable, ongoing standard. What does that mean for high…
24/08/2026
Infrastructure
All Infrastructure →
Model Router spreads requests across AI providers
Microsoft's Model Router routes across several AI providers with automatic fallback, but reduces concentration risk only with an i…
28/10/2026
Verifying routing, trust anchors and robot control before you call an AI workflow reliable
RPKI updates from APNIC and RIPE NCC and fresh robot vulnerabilities show why an AI workflow is only reliable once you can verify …
16/09/2026
Exit and fallback for AI services: what the July 2026 OpenAI outage exposes
After the 17-day OpenAI outage of July 2026, exit rights, data portability and tested failover prove to be a governance layer, not…
25/08/2026
Authors
All authors →Marit Halversen
Covers AI governance and regulatory design, with a focus on how compliance obligations land on architecture ra…
Tobias Lindqvist
Adversarial machine learning and the security properties of retrieval systems.
Noor El Amrani
Data protection, anonymisation practice, and what regulators actually accept as evidence.
Follow by feed
No newsletter and no account — the site sends no mail at all, and the feed carries no analytics.