👋🏻 Hello! We’re now 3,190+ reading ”Sensible AI” to move from headlines to decisions, in minutes.

This week, coordination and controls are rising: the G7 aligns positions, the UK proposes tough rules for frontier models, and China tightens labeling and suggests an ASEAN framework.

We’ll close with your “from the news to your action plan” section so you can act now!

Let’s begin 👇

^{Spanish version here: IA con Sentido – Español}

TOP NEWS

🌍 G7 meets in Montreal on AI
→ Ministers from the seven countries gather in Canada to coordinate positions on algorithmic governance and quantum computing.

🇬🇧 UK to regulate powerful AI
→ 100+ lawmakers call for the most advanced foundation models to face strict legal obligations for national security.

🌍 UK survey shows distrust in AI
→ A 25-country study finds most people fear harmful uses and demand independent oversight of companies and governments.

🇨🇳 China tightens labeling, proposes ASEAN framework
→ The cyber regulator fines services for not flagging synthetic content and proposes shared ASEAN rules on data and algorithms.

WHAT HAPPENED WORLDWIDE?

🇪🇺 EU flags low readiness and AI Act gaps
→ European agencies warn many entities skip human-rights assessments and could sidestep high-risk rules in migration.

🇺🇸 Washington state sites with illicit links
→ State government portals unknowingly hosted external links to synthetic sexual images created by generative models.

🇪🇸 Spain modernizes services with Mistral
→ Central government and Madrid agree to use European models to speed public services while maintaining data protection and transparency.

🇺🇸 U.S. agencies publish AI plans
→ Federal departments present strategies to automate processes while protecting civil rights and impose new demands on tech vendors.

🤖 Boomi: data-center governance guide
→ The integration provider explains how to register models and flows to meet regulatory frameworks from enterprise infrastructure.

🇬🇭 Ghana activates UNESCO AI assessment
→ The communications ministry begins a national ethics-readiness diagnosis to guide public policy and a future tech strategy.

🌍 Alliance InclusiveAI global coalition
→ Academic and civil organizations launch an open alliance to promote inclusive systems that share AI benefits at scale.

🇨🇱🇲🇽 Chile–Mexico pact on responsible AI
→ Both governments sign bilateral cooperation for research, regulation, and joint projects to drive innovation without widening regional gaps.

🏫 Google’s Nordic education alliances
→ Agreements with schools in Finland, Sweden, and Norway for pilots combining digital personalization and student privacy.

🇪🇬 Egypt schools pilot ethical AI
→ Schools integrate assistants and supervised social platforms to improve focus, well-being, and teen engagement.

🇪🇸 Spain will host a military-AI summit
→ Foreign Affairs announces an international meeting in 2026 to set principles on autonomous weapons and defensive uses of intelligent systems.

AND AT ENTERPRISE LEVEL?

🇺🇸 Perplexity faces new copyright suits
→ The Chicago Tribune and The New York Times sued Perplexity for allegedly copying articles without permission and generating infringing content.

🤝 Microsoft: clients demand ethical development
→ A company executive says large enterprises condition contracts on assurances about safety, transparency, and limited use.

📣 SAP’s CEO urges lighter EU rules
→ The German executive calls for simpler regulations to compete with the U.S. and China in enterprise software powered by advanced models.

🏛️ Deloitte on internal audit’s AI role
→ The firm details how to review data, models, and change controls to reduce legal and operational risk.

🧾 ImpactoTIC algorithm audit checklist
→ A Colombian outlet offers a practical list for executives to review governance, model inventories, and bias controls.

⚙️ CSOOnline warns of AI governance debt
→ Delaying clear policies will create costly future failures in cybersecurity and compliance.

🐾 AVMA’s AI framework for veterinary medicine
→ The association sets principles for clinical algorithms that support diagnosis while keeping final decisions with human professionals.

🤝 Forbes: brands build ethical trust
→ A report shows companies prioritize transparency, plain-language explanations, and user control to strengthen customer relationships.

⚖️ FCPA: guidance on AI in compliance
→ Experts recommend supervised systems that detect bribery and anomalous reports without discrimination or privacy violations.

🧠 Microsoft Spain stresses human judgment
→ The innovation lead notes statistical models lack common sense; critical decisions must remain human.

📈 EY shows the business value of responsible AI
→ Clear policies help measure ROI, reduce incidents, and convince regulators and boards.

ORGANIZATIONS & STANDARDS

🇺🇳 UN warns of a new AI divide
→ Report says wealthy countries concentrate capabilities while developing economies risk dependence on external platforms.

📜 UNESCO proposes practical ethical anchors for AI
→ The organization outlines steps to turn principles into committees, indicators, training, and mandatory periodic assessments.

FROM THE NEWS TO YOUR ACTION PLAN

I turn these headlines into concrete steps you can start today to strengthen your AI governance. Pick 1 action to get results under 7 days:

🟪 Implement mandatory labeling of synthetic content → Ensure visible notices on AI-generated text and images, aligned with new Chinese rules.
🟪 Activate human oversight for critical decisions → Keep human review in automated processes, as urged by Microsoft Spain and the AVMA.
🟪 Establish a living inventory of models and data → Register models, versions, and sources as recommended by Deloitte and Boomi.
🟪 Review rights and privacy in public-sector automation → Adopt controls like those required by U.S. federal agencies to protect civil rights.
🟪 Create an operational ethics committee with KPIs → Turn principles into measurable processes following UNESCO’s practical anchors.
🟪 Build contract-level responsible-use policies → Bake in the guarantees major Microsoft clients require in your vendor agreements.

That’s all for today!

What topic would you like me to dig into on Friday? Your reply guides the next edition.

If this helps your team, forward it and tell me what worked: I improve the newsletter with your feedback 💌