Here’s what caught my eye today across AI policy, small business adoption, marketing, and music production. Ten stories, no fluff, with a few thoughts on why each one matters for the work we do.
1. OpenAI and Anthropic’s CEOs told the UN they can’t be trusted to govern AI alone
Sam Altman told the UN Security Council that humanity could lose control of AI’s future if the biggest decisions stay locked inside a handful of San Francisco labs. Dario Amodei went further, saying poorly managed AI could be a risk to humanity as a whole. The Trump administration flatly rejected the idea of international oversight in the same session. I don’t think either extreme gets this right, unchecked global bureaucracy or zero guardrails, but when the people building the technology are the ones asking for rules, that’s worth paying attention to.
2. Trump tells Xi no guardrails are coming on AI, even as industry leaders warn of losing control
A day after AI executives asked the UN for shared rules of the road, the US signaled it has no interest in coordinating with China on AI safety either. Two countries racing each other with no agreed limits is exactly the scenario people have been warning about for years. For those of us doing this work at the ground level, the lesson is the same either way: the guardrails aren’t coming from the top, so we build them into how we deploy AI with our own clients.
3. AI data centers are reshaping rural land, and rural communities are starting to push back
Two thirds of planned US data centers are headed for rural areas, not cities. That means jobs and tax revenue, but also noise from cooling systems, strain on water and electric grids, and zoning fights most small towns have never had to think through before. If you’re in a rural community watching this land up close, this is worth reading before a rezoning hearing shows up on your town’s agenda.
4. A bipartisan bill would have the federal government set real guardrails on autonomous AI agents
Reps. Mike Lawler and Josh Gottheimer introduced the Stop Rogue AI Act, which would have NIST build guidelines for identifying and controlling AI agents before federal agencies and contractors deploy them. It’s a narrow bill and it won’t make many headlines, but this is the kind of practical governance work that matters more than most of what does. Autonomous agents are already showing up in small business tools. This is the groundwork for keeping them accountable.
5. The AI for Main Street Act passed the House, and it’s aimed straight at businesses like the ones I work with
This bill directs Small Business Development Centers to actually train small business owners on evaluating and adopting AI responsibly, not just hand them a chatbot and walk away. It passed the House with real bipartisan support, which tells you something. If it becomes law, it means federal backing for exactly the kind of hands-on, ethics-first AI adoption work I’ve been doing on my own.
6. Goldman Sachs: 76% of small businesses use AI, but only 14% have it built into how they actually operate
That gap is the whole ballgame. Most owners have tried an AI tool somewhere, but almost none have integrated it into daily operations in a way that changes the business. 73% say they want more training and support to get there. That’s not a technology problem, that’s a consulting and change-management problem, and it’s exactly the gap this business exists to close.
7. Amazon rolled out “always-on” AI workflows for sellers, with a Claude-powered plugin
Amazon’s Seller Assistant now handles pricing and restocking decisions continuously instead of waiting to be asked, and it’s already reaching over 90% of sellers worldwide with high acceptance on its recommendations. For small sellers this is a preview of where a lot of business software is headed: less “ask the AI a question” and more “the AI is already working in the background.” Worth understanding before that becomes the default everywhere.
8. A Marketo co-founder just launched marketing automation built to reason, not just follow rules
Jon Miller’s new platform, Phave, uses reasoning models to make marketing decisions inside guardrails the business sets, instead of old if-this-then-that automation. It’s priced for mid-size operations right now, not a solo shop, but it signals where marketing tools are heading. The guardrails piece is what’s worth watching. That’s the difference between AI that helps and AI that runs off without you.
9. The Suno and Udio lawsuits are settling out, and streaming platforms are quietly deciding how AI music gets labeled
Warner settled with Suno, Universal settled with Udio, and both are building licensed AI models going forward. Sony is still fighting it out in court, with rulings expected this summer. Meanwhile Deezer, Spotify, and Apple Music are each taking different approaches to flagging AI-generated tracks, none of them fully public about it. If you’re producing music with AI tools, the honest move is the hybrid one most working musicians are already landing on: AI for demos and sketches, real production for what you actually release.
10. Google’s AI music tool, Flow Music, added the ability to remix and extend your own tracks
Formerly ProducerAI, now folded into Google’s Flow product family and built on Gemini and Lyria 3. The new features let you take a generated track and extend a section or try variations until it fits, which moves it closer to an actual production tool instead of a one-shot generator. Still not a replacement for a producer’s ear, but it’s closing that gap faster than most people expect.
That’s the digest for today. If any of these touch something you’re wrestling with in your own business or ministry, reach out. That’s exactly the kind of conversation this newsletter exists to start.