Workday lawsuit, DORA logs, and *deepfake evidence* on trial
A discrimination case asks how AI hiring tools make decisions, financial logs must resist tampering, and courts face AI-generated content.
- 01
Workday sued over AI recruitment discrimination
Derek Mobley alleges Workday's AI recruitment software discriminates by race, age and disability, rejecting him from over 100 applications. The case raises whether the company can prove how its AI made hiring decisions.
Why it matters for evidence: Article 12 requires high-risk recruitment AI to log events for traceability. If Workday can't show how decisions were made, the defence gets harder.
Source: facebook.com
- 02
AI-generated content challenges evidence authenticity
OpenAI and Meta told a US federal judicial panel that AI audio and video are increasingly hard to distinguish from authentic content, raising challenges for evidence authenticity and organisations' ability to prove what their AI actually did.
Why it matters for evidence: Swedish courts assess evidence freely. If you can't show what your AI produced or whether content is authentic, the other side's version may prevail.
Source: linkedin.com
- 03
Academy explains DORA's five pillars for finance
An academy outlines DORA's digital operational resilience pillars, including ICT third-party risk and ISO standards. DORA requires financial entities to protect logs against manipulation and unauthorised access.
Why it matters for evidence: Article 12(2)(d) of the DORA technical standard says logs must be protected against tampering. How you prove they weren't changed is the evidence question.
Source: standards.academy
- 04
Regulations.md publishes AI Act overview
The site summarises the EU AI Act, adopted March 2024 and entering force 1 August 2024. It covers transparency, logging, data governance, human oversight and impact assessments for high-risk systems.
Source: regulations.md (in Swedish)
- 05
Austrian marketers prepare for AI labelling rules
The Austrian Direct Marketing Association is holding an event on Article 50 of the AI Act, which from 2 August 2026 requires disclosure when AI generates images, video or audio that appear authentic.
Source: werbungwien.at (in German)
- 06
Europe Cyber-Safe launches AI Trust Passport
Europe Cyber-Safe offers a manually reviewed evidence package for AI suppliers selling to regulated buyers in Europe. The concept is an open disclosure rather than a certificate, designed to help suppliers show compliance with documentation, logging and traceability requirements.
Source: cyber-safe.eu (in Swedish)