Weekly AI Briefing: July 28, 2026

Three stories from the past week, with context you can use quickly.

1. Europe’s AI transparency rules arrive on August 2

What changed: The EU AI Act transparency obligations apply from 2 Aug 2026.
Official basis: Article 113 (application date) and Article 50 (transparency duties), plus July 2026 Commission guidance.
What users should expect: More explicit labels such as “AI-generated” and clearer non-human chatbot disclosures.
Why this matters globally: Large AI providers often standardize behavior across regions, not only EU markets.

Federal reality: There is still no single US federal chatbot law in force.
State reality: The FPF tracker shows ~100 chatbot-specific bills across 34 states.
Common policy direction: Non-human disclosure, minor safeguards, and tighter data-use limits.
Markey bill status: S.4199 Youth AI Privacy Act is introduced and referred to committee, not enacted.
Practical takeaway: For families with teenagers, protections depend heavily on your state, not one national standard.

Dataset size: Google analyzed 15 million anonymized conversations in ATLAS.
Main signal: AI is being used more as assistive support than full job replacement.
Notable detail: Use appears across unexpected roles, including hands-on technical work.
Important caveat: This is a company-published dataset, so treat it as informative, not definitive.
Reader takeaway: Avoid extreme narratives. Track where AI adds value while preserving human judgement.

Have a story you think belongs in next week’s briefing, or a question about something you read here? Reply to this newsletter or visit Contact.