NASA’s open-source lunar AI model lands, California drops the nation’s toughest kids’ chatbot rules, and Qatar funnels $875M into AI chips—while the UK Greens move to ban killer robots and Malaysia rewrites its AI governance playbook. Scroll down to catch the signals that matter.

🌙 NASA-IBM Lunar AI Model Rewrites Moon Science

NASA and IBM’s Lunar Foundation Model turns millions of high‑resolution and multispectral Moon images into a single geospatial AI engine for mapping craters, polar ice and volcanic terrain. Built on decades of Lunar Reconnaissance Orbiter data and released open-source, it gives researchers reusable code, datasets and benchmarks for new lunar missions.

Expect faster, AI-assisted lunar discovery pipelines and more data-driven site selection for future Moon bases.

🛡️ California’s Kids’ AI Safety Package Sets a New Bar

The Transparency Coalition spotlights California’s sweeping child-tech package, where Governor Newsom signs 13 laws strengthening safeguards for companion chatbots and social platforms used by minors, including Adam’s Law SB 1119 and new duties around crisis protocols and manipulative outputs, detailed in their kids’ safety brief.

Children’s AI and social media experiences will now be designed under strict legal obligations, not just voluntary trust-and-safety policies.

👧 SFGATE: Newsom Tightens AI Chatbots for Children

SFGATE reports that Gov. Gavin Newsom, flanked by parents and lawmakers, signed 13 child-safety bills targeting social media algorithms and AI chatbots used by kids, including measures limiting “addictive features” for users under 16 and expanding liability for harm, as outlined in their Bay Area coverage.

Platform design for minors in California now faces direct legal scrutiny, making kid-safe interaction patterns a regulatory requirement.

📜 Malaysia Completes Draft AI Governance Bill

Malaysia’s digital minister confirms the country’s first dedicated AI Governance Bill is complete and ready for Cabinet review, with plans to table it to Parliament next. The Malay Mail notes a risk-based framework aimed at deepfakes, content integrity and lifecycle-wide oversight aligned with global best practice.

Malaysia is positioning AI rules as a central pillar of its digital strategy, signaling tighter compliance expectations for regional AI deployments.

🏃‍♂️ Frontiers Maps Ethical Risks of AI in Physical Education

A systematic review in Frontiers in Public Health examines how AI in physical education can generate privacy leaks, algorithmic bias, homogenized teaching and body-related harms, while outlining governance strategies like explainable systems and strengthened data security, detailed in the article on ethical risks in AI PE.

Education and sports programs using AI will increasingly need discipline-specific ethics frameworks, not generic edtech guidelines.

🏭 NVIDIA & Palantir Launch Sovereign AI for Supply Chains

NVIDIA and Palantir unveil a sovereign AI stack that fuses Nemotron open models with Palantir Foundry and AIP to orchestrate complex supply chains, first inside NVIDIA’s own operations. Their joint announcement highlights a reference architecture for on-prem or cloud, enabling governments and manufacturers to retain data control.

Critical infrastructure and chip supply chains are becoming AI-orchestrated systems where sovereignty and hardware choice are strategic levers.

🤝 Salesforce Completes $3.6B Fin AI Agent Deal

Salesforce has closed its roughly $3.6 billion acquisition of Fin (formerly Intercom), folding its autonomous AI customer agents into Salesforce AI Labs and CRM, as broken down in this completion analysis from Vantagepoint. Fin’s Apex-powered agents resolve complex queries end-to-end across chat, email, messaging apps and voice channels.

Customer service is shifting toward packaged, agentic AI stacks, with Fin giving Salesforce a turnkey autonomous support layer for 30,000+ customers.

🧭 UB Researchers: Responsible AI Must Span the Entire Lifecycle

University at Buffalo researchers argue responsible AI must be embedded from data collection and model design through optimization and governance, not bolted on after deployment. Their piece on responsible AI for social good emphasizes social objectives, harms analysis and institutional accountability as core design constraints.

Expect growing pressure on AI teams to treat ethics, fairness and social impact as first-class engineering requirements, not post-hoc compliance.

💸 QIA Backs Positron AI’s $875M Series C

Qatar Investment Authority announces participation in Positron AI’s USD 875 million Series C round, valuing the AI inference hardware company at USD 5 billion, according to the sovereign fund’s QNA dispatch. The capital supports Asimov chip design and scaling Titan high-memory inference systems.

Large sovereign capital is betting heavily on specialized inference silicon, intensifying competitive pressure on incumbent AI hardware players.

🌱 UK Green Party’s Five-Point Plan Targets AI Risks

The UK Green Party unveils a five-point AI plan calling for a strict regulatory system, a global oversight body, bans on lethal autonomous weapons and harmful manipulative uses, plus worker protections and public-good funding, detailed in the party’s policy blueprint. It frames AI as a potential existential risk requiring precautionary regulation.

Political pressure is mounting for binding, global AI rules that explicitly restrict military and manipulative applications, not just commercial misuse.

Which of today's stories changes your roadmap? Tell us in one line.

Forward the signal: share your referral link