Big Creek Briefing

AI is Working For Small Business

Welcome to this week's Big Creek Briefing.

This week felt like a turning point. Between Anthropic plugging Claude directly into the tools you already run your business on, OpenAI opening its ad platform to companies that can't afford a $50K minimum spend, and a fresh wave of data showing AI is now splitting founders into two distinct camps (growers and decliners), the window between "early adopter" and "left behind" is closing fast. Here is what you need to know heading into the week.

Let's jump in!

AI Joke of the Week

My accountant told me I should start using AI to cut costs. Claude told me I should replace my accountant. They're both right. Neither one is happy about it.

Big Stories of the Week

Anthropic Launches Claude for Small Business: Anthropic officially launched Claude for Small Business, a package of connectors and ready-to-run workflows built inside Claude Cowork. It integrates directly with QuickBooks, PayPal, HubSpot, Canva, Docusign, Google Workspace, and Microsoft 365, targeting the highest-friction tasks small business owners face every day: month-end close, invoice chasing, payroll cash-position forecasting, and contract handling. Every action requires user approval before executing, a deliberate design choice aimed at founders who want the power of AI agents without handing over the keys entirely.
OpenAI Opens ChatGPT Ads to Small Business:
OpenAI launched a self-serve Ads Manager for ChatGPT, and the headline that matters most to small business owners is this: the previous $50,000 minimum spend requirement is gone. Any business can now run ads that appear as clearly labeled, contextually matched placements at the bottom of ChatGPT responses. The platform uses conversation-based targeting rather than traditional keyword matching, which means your ad shows up when someone is literally in the middle of a relevant conversation. This is a new channel worth watching closely.
OpenAI Creates the OpenAI Deployment Company: OpenAI announced the formation of the OpenAI Deployment Company, a separate entity backed by more than $4 billion in initial investment, designed to embed engineering teams directly inside organizations to accelerate real AI adoption. The company acquired Tomoro to bring Forward Deployed Engineers into the mix. For founders, the signal here is clear: the era of "we'll figure out AI eventually" is over. The biggest players in the space are now sending people into businesses to make it happen.

Main Insights

AI Adoption Numbers are Impossible to Ignore

The data this week makes the business case for AI adoption impossible to ignore. According to a 2026 SBE Council survey, 68% of U.S. small businesses now use AI regularly, up from 48% just a year ago, and companies deploying AI in core functions are reporting 26 to 55% productivity gains. The number that stands out most: 91% of small business AI users say it is boosting their revenue, and businesses using AI are nearly twice as likely to report year-over-year revenue growth compared to those that are not. The practical takeaway is that the typical AI-using small business now runs a stack of five tools embedded into daily operations, meaning the question is no longer whether to adopt AI but how fast you can build the right stack.

Founder Insights

Here is the split that should get your attention: 83% of growing small businesses have adopted AI, compared to just 55% of declining ones. That gap is not about budget, it is about behavior. Founders who are growing are treating AI as infrastructure, not a curiosity. Meanwhile, overall deep adoption (AI embedded into core workflows, not just used occasionally) still sits at only 7% of small businesses, which means there is a wide open window right now for founders willing to go beyond surface-level use. The founders who build an actual AI operating layer in the next 12 months will have a structural advantage that is very hard for competitors to reverse.

Quick Hit: One Prompt to Simplify Your Week

"I run a [type of business] with [X] employees. My three biggest operational time drains right now are [list them]. For each one, suggest a specific AI tool or workflow I could implement this week to cut the time in half. Be specific about the tool, the setup steps, and what I should realistically expect."

This prompt forces your AI assistant to think strategically about your specific business instead of giving generic advice. Run it monthly and compare the outputs.

AI Tool Spotlight: Osaurus

What it does: Osaurus is a new Mac-native AI tool (covered by TechCrunch this week) that lets you run both local and cloud AI models from a single interface, entirely on your own hardware. Think of it as a personal AI control center that can access your files, browse your apps, and manage system tasks, all without sending sensitive data to a third-party server. You choose whether a given query runs locally on your machine or gets routed to a cloud model, giving you real control over privacy and performance depending on the task.

Why founders should care: If you are running any kind of client-facing operation, handling financial data, or working on proprietary content, the ability to keep sensitive work off cloud servers is genuinely valuable. Osaurus makes it practical: you can handle confidential internal drafts and strategy documents locally, then flip to a cloud model for general tasks. For founders at agencies, in legal-adjacent fields, or anyone who signs NDAs, this kind of setup is increasingly worth having. It is also open source, meaning the community can audit exactly what it does with your data.

Pricing and availability: Osaurus is free and open source, available now for Apple Silicon Macs. You can find it on GitHub and the project is currently being incubated through the Alliance accelerator in New York. Search "Osaurus AI Mac" or head to TechCrunch's May 15 write-up for a full walkthrough and download link.

What We're Watching

Government pre-release AI testing is becoming standard

The U.S. government is now requiring major AI labs, including Microsoft and xAI, to provide early model access to regulators before public launch. This signals a meaningful shift in how AI products will reach market and could affect release timelines for tools you rely on.

Multi-agent AI architectures are moving out of the lab

Enterprises are starting to deploy systems where multiple AI agents collaborate on complex workflows, each handling a different piece and passing context between them. IDC projects AI copilots will be embedded in nearly 80% of enterprise workplace applications by end of year. The small business version of this is coming fast.

AI is starting to build AI

TechCrunch published a notable piece this week on what happens when AI systems begin generating and improving their own models. This is early but the acceleration implications for the tools founders use are significant

The AI adoption gap is a competitive moat in the making

With deep AI adoption still under 10% among small businesses, founders who move now are not just saving time. They are building a gap that compounds. Watch for more data on how the top-performing small businesses are structuring their AI stacks.

The Wrap(TLDR)

Anthropic and OpenAI both made major moves for small business this week, dropping the barriers to AI-powered operations and advertising significantly. The data is unambiguous: AI-adopting businesses are growing faster, and the gap between them and non-adopters is widening. Deep adoption is still rare, which means right now is your window. Find the two or three highest-friction tasks in your business and automate them this month.

If you want help designing AI workflows or understanding how these new models fit into your business, Big Creek Growth can help. Reply to this email and we will walk through it together.