AI This Week: Agents Take the Wheel
This week, Microsoft officially called it: we’ve entered the “agentic era” of AI. Meanwhile, Colorado softened its landmark AI law before it even took effect, and fresh data shows small business AI adoption just crossed 58%. A lot moved in seven days.
Microsoft Declares the Agentic Era — And SMBs Are in the Crosshairs
At Build 2026 (June 2–5 in San Francisco), CEO Satya Nadella announced that Microsoft is done with AI that just assists — the company is now building AI that acts. The new Copilot Platform gives businesses a set of APIs and connectors that embed AI agents directly into existing workflows. If you’re running Microsoft 365, your tools are about to get a lot more proactive.
Two features stand out for small and mid-size teams. Work IQ maps how work actually flows across your Microsoft 365 apps and organizational systems, then grounds agents in that context so they can take action — not just answer questions. Frontier Tuning (currently in private preview) lets agents learn the specific way your business operates, within your compliance boundaries.
This isn’t “AI as a chatbot you open in a new tab.” It’s AI woven into the apps your team already uses — Outlook, Teams, Excel — handling the follow-up tasks, scheduling, and document work you currently do manually.
What this means for your business: If you’re a Microsoft 365 shop, you don’t need a big technology overhaul to start using AI agents. The capability is coming to you. The question is whether your workflows are documented clearly enough for agents to actually help — most SMBs aren’t there yet, and that gap is worth closing now.
Microsoft’s New AI Models Cost Up to 10x Less Than the Competition
Alongside the agent announcements, Microsoft launched seven in-house AI models — the MAI family — that it says deliver enterprise-grade performance at a fraction of current market prices. The headline model, MAI-Thinking-1, is Microsoft’s first reasoning model. Another, MAI-Code-1-Flash, generates source code from plain-language descriptions.
The cost comparison is striking: Microsoft says MAI models matched GPT performance on enterprise tasks while running at up to 10x lower cost. These models are available now through OpenRouter, Fireworks, and Baseten — not locked inside Azure.
For businesses using API-based AI services, this matters beyond Microsoft specifically. More competition in the foundation model market pushes prices down across the board, including from OpenAI and Anthropic.
What this means for your business: If cost has been your reason for holding off on AI tools, this week’s news changes the math. The 61% of SMBs that cite cost as their top adoption barrier now have more options at the affordable end of the market.
Colorado’s Sweeping AI Law Got a Business-Friendly Rewrite
Colorado’s original AI Act (SB 24-205) was set to take effect June 30 — and it was strict. Businesses deploying “high-risk AI systems” in decisions about employment, housing, or credit would have faced mandatory risk management programs, impact assessments, and potential fines up to $20,000 per violation.
Then, on May 14, Governor Jared Polis signed SB 26-189, which repeals and replaces the original law entirely. The new framework is disclosure-based: if AI influences a consequential decision affecting a Colorado consumer, you disclose it. The duty-of-care requirements, risk management mandates, and impact assessments are gone. The U.S. Chamber of Commerce had flagged the original law as a threat to small business AI adoption — the replacement is a significantly lighter lift.
Don’t confuse Colorado with the full picture, though. Illinois HB 3773 still requires notification when AI assists with hiring, promotions, or performance reviews. And the EU AI Act compliance deadline for certain high-risk systems hits August 2 — relevant if you have customers in Europe.
What this means for your business: If you use AI to screen job applicants, evaluate credit, or assist with healthcare decisions, you need a disclosure policy — even under the simplified law. If you operate in multiple states or have European customers, get a one-page compliance checklist together before August. It’s a quick win that protects you from a costly oversight.
New Data: 58% of Small Businesses Now Use AI — With 91% Reporting Revenue Gains
Research published this week shows small business AI adoption jumped from 40% to 58% in a single year. That’s a 45% increase — faster growth than mobile adoption saw at a comparable stage.
The results for early adopters are hard to ignore: 91% of SMBs using AI report revenue increases, and AI-assisted customer service teams report 23% higher customer satisfaction scores. These aren’t enterprise numbers — they’re coming from small businesses. Houston operators in professional services, healthcare, and logistics are seeing the same pattern locally.
The uncomfortable flip side: only 12% of SMBs have a dedicated AI strategy. Most are using individual tools — a chatbot here, an image generator there — without a coherent plan. That’s fine for getting started, but it explains why so many businesses use AI without seeing the outsized results others report.
What this means for your business: You’re probably already using some AI. The question is whether you’re using it strategically. Businesses that have mapped their workflows, identified their highest-leverage use cases, and trained their teams consistently are the ones hitting those revenue numbers. Random tool adoption doesn’t get you there.
The takeaway
The thread running through this week’s news: AI is getting cheaper, more embedded in existing tools, and more regulated in specific use cases — all at the same time. That combination actually makes the path forward clearer, not murkier. Lower costs reduce the barrier to entry. Embedded agents mean less change management. Narrower, disclosure-based regulation is something any business can handle.
If you’re a Houston business owner figuring out where to start — or where to go next — we can help you build a practical AI strategy that fits your workflows, your team size, and your budget. No buzzwords, no bloated projects.