Temper news

AI governance changes. Your evidence should keep up.

Short, source-linked briefings on the changes that can alter an AI claim, a buyer answer, or a compliance plan.

This is the operating layer of a live trust profile: we identify the change, explain the exposure, and show what should be re-verified.

Latest signal

4 September 2026
European Union · Digital Services Act

EU designates ChatGPT a “Very Large Online Search Engine”

The Commission has put ChatGPT under the DSA’s VLOSE regime after the service reported at least 45 million average monthly EU users. The designation is about this service and its web-search role—not a blanket ruling that all AI is a search engine.

What changed

A new regulatory category applies

ChatGPT must meet the DSA’s additional requirements for very large online search engines, with compliance due by January 2027.

Why it matters

AI answers are now a regulated distribution surface

The focus shifts beyond model behavior alone to systemic risk from the service and its algorithms: illegal content, minors, wellbeing, fundamental rights, elections, and public security.

What to check

Revisit claims tied to public-facing AI

  • Does your risk register cover search and answer-distribution behavior?
  • Do safety, transparency, and incident evidence identify the deployed service—not only the model?
  • Could customer-facing compliance answers need a new qualification?
Read the Commission release ↗

Not a headline feed

Temper follows the primary record. Each briefing states what is known, what it changes, and which evidence may be affected. The Regulatory Tracker remains the source-level map; this page is the customer-facing interpretation layer. Informational only—not legal advice.