// NOTES
Building in public
Field notes from Jane Riveros — fifteen years of enterprise program, risk, and resilience leadership in regulated financial services, now building governed AI agents. Essays on what regulated delivery teaches you about deploying agents that can't be allowed to improvise.
Why I build AI agents like regulated programs
MAVIS and Victoria aren't demos — they're production agents built with governance gates, audit trails, and permission-scoped access. Here's the discipline I carried from the bank to the build.
Read the manifesto →Chile's AI ambition needs more than a headline
Chile wants to become a regional tech hub through AI alliances, but without a governance backbone that is just wishful thinking. I have seen too many countries chase the shiny object without building the ethical and compliance infrastructure to sustain it.
Don't Let Agentic AI Hype Outrun Your Governance
Gartner just named agentic and physical AI the top supply chain trends for 2026, but most organizations are nowhere near ready to govern them. If you think today's AI compliance is hard, wait until your bots start negotiating contracts without telling you.
Spain's AI sandbox is a smart start for biometric rules
Spain is launching an AI sandbox to test biometric systems against the EU AI Act before enforcement. This is the kind of hands-on regulatory experimentation we need, not more theoretical white papers.
Partial Approval for Mythos 5 Sends a Mixed Signal
Anthropic won partial U.S. approval for Mythos 5 as regulatory pressure on AI increases. We need clearer transparency and governance frameworks.
Invited, Authorized, and Out of Control: Re-evaluating Trust for Agentic AI
Agentic AI is changing how I think about cybersecurity. The biggest threat is no longer unauthorized access anymore but what happens when autonomous agents are trusted with “too much access.”
State AI regulation is here. Now we need to get it right.
A news roundup on state AI regulation highlights the growing patchwork of rules that governance teams must navigate. Compliance is not a one-size-fits-all exercise, and ethics must be woven into every implementation decision.
Tax compliance is the canary in the AI coal mine
Manufacturers rushing to deploy AI often overlook tax implications until audit season. Wolters Kluwer shows that responsible AI research can surface impacts early, but the real lesson is for governance teams everywhere.
When Bannon and Sanders agree on AI regulation, pay attention
Steve Bannon and Bernie Sanders share almost nothing politically, but both call for AI regulation. That rare alignment is a signal that governance practitioners need to stop waiting and start building measured compliance frameworks now.