China’s AI agent framework took effect on July 15, 2026, introducing tiered authorization, human oversight and requirements for certain high-risk deployments.

This issue brief argues that the companies behind consumer-facing AI agents should owe their users a duty of loyalty.

A policy report by Juan David Gutiérrez and Sebastián Hurtado, dated August 2026, summarises the first webinar of the series on AI governance in Latin America, held on July 15, 2026.

The rapporteur said Brazil’s AI regulatory bill is now expected to be voted on late in the year, after the October elections.

CIPESA recommends that Kenya align its AI proposals and add enforceable rights, oversight, participation and environmental safeguards.

Georgetown's CSET announced new university partners to grow its AGORA database of over 1,000 AI laws, regulations and standards.

The 120-company Open Secure AI Alliance, spearheaded by Nvidia, released initial proposals via its SAFE working group for reporting AI cybersecurity incidents.

The Commission issued guidance clarifying Article 50 transparency duties for AI providers and deployers ahead of the rules taking effect in August 2026.

An Al Jazeera explainer published August 6, 2026 examines how the EU AI Act's transparency rules took effect while high-risk obligations were postponed via the Digital Omnibus.

Microsoft's industry-backed "Open Weights and American AI Leadership" letter reached more than 270 company and organization signatories as of August 3, 2026.

OpenAI identified on August 7 that its Astra model could have critical-level cyber capability, triggering stricter chain-of-thought monitoring for all inference with tools.