Newsletter

You're Trusting AI With Your Insurance. That's a Problem.

Scott McIntoshAugust 20, 20264 min read
Featured image for You're Trusting AI With Your Insurance. That's a Problem. - Newsletter article by DigitalTreehouse

Plus: Meta just gave you a free analyst for your ad account, and Claude is moving into your inbox.

Meta built an analyst into your ad account. If you're running Facebook or Instagram ads, Meta quietly did something useful this week. The new AI capabilities inside Meta AI connect directly to your actual account data, not generic marketing platitudes. You can ask it to break down campaign performance, see how you stack up against competitors, generate reports, and set up recurring tasks, all without leaving the platform. The competitive benchmarking piece is the one worth paying attention to. What used to take a spreadsheet, a few browser tabs, and a solid chunk of a Tuesday morning can now happen in a question or two. This is the version of "AI for marketing" that's actually practical. Read more.

Claude is showing up where you already work. Anthropic had a busy week. Nothing flashy, but the pattern is clear. Claude can now draft and send Gmail replies, manage files in Google Drive, and sync your sessions across your phone and desktop. Their Claude Cowork product is now on mobile for paid users, and they bumped Claude Code usage limits 50% through August. Taken one at a time, none of these is a big deal. Taken together, Anthropic is trying to turn Claude into the assistant that's already open in the background, not one more tab you have to remember to switch to. If you're on a paid plan and you live in Gmail and Drive, it's worth 20 minutes to see what it can actually do for you now. Read more.

Owners trust AI for insurance advice. Most are still underprotected. A survey of 501 US small business owners by ERGO NEXT Insurance turned up something worth sitting with. Thirty-six percent now use AI chatbots or search tools to research business insurance. Of those, 94% trust what they got back. And 31% trust AI more than a human agent. Here's the problem: 73% of small businesses still carry inadequate coverage. AI gives confident, polished answers about insurance. It doesn't know your lease terms, your payroll, your equipment, or the specific liability exposure in your industry. It's like asking a well-read friend who's never run a business to explain your risk profile. Useful starting point. Dangerous endpoint. If AI is how you last researched your coverage, use it to generate a list of questions, then run those questions by an actual broker. Read more.

Tools Worth a Look

Calendly Notetaker + Callie: Two new products from Calendly. Notetaker captures meeting notes and action items automatically; Callie handles prep and follow-up so you can stay present. Good for owners who lose time after every client call.

Firefox Smart Window: An AI layer inside Firefox that answers questions with live web data, groups your open tabs, and keeps everything stored locally on your machine. Free, and nothing goes to the cloud.

SureThing: An AI social media agent that learns your voice and posts to LinkedIn, X, Instagram, and TikTok on its own. For owners who want a social presence without the daily effort.

Khan TED Institute: Sal Khan's planned AI-era college, a full undergraduate degree for under $10,000, built with Google, Microsoft, McKinsey, and Replit. Not yet accredited, launching 2027. Worth watching if you're thinking about affordable AI education for employees.

More Headlines

What small businesses actually want from their bank's AI: New Stax research says owners are open to AI-powered banking tools but won't trust ones that don't connect to their actual account data, generic financial chatbots are a hard sell.

Congress has an AI slop problem, and it's a cautionary tale for any owner who publishes AI-drafted documents: The House is getting flooded with AI-written bills full of miscited legal codes and wrong jurisdictions, and one official says fixing the drafts takes longer than starting from scratch.

Critics insist AI will ruin us. These small business owners and workers already know better.: Real owners share what's actually changed in their day-to-day, worth a read if you're still on the fence.

India's small businesses could be AI's biggest winners, according to a new World Bank report: The argument is that AI levels the playing field between small operators and large competitors more than any previous technology wave.

Stripe pays $7B for OpenRouter: OpenRouter lets businesses route tasks across multiple AI models. Stripe's bet is that this kind of infrastructure becomes essential plumbing as AI use grows.

OpenAI pauses frontier training after a safety containment failure: They've added 20% compute overhead for monitoring. A reminder that the frontier labs are still working out real safety problems, not just PR ones.

AI watermark removal tools are proliferating: Anthropic's Claude watermarking effort has already spawned countermeasures. The cat-and-mouse dynamic around AI-generated content authentication is just getting started.

Practical framing on AI adoption for small businesses from Graham Peddie, founder of Frame and Co.: A founder-level perspective on where to start and what to ignore.

-Scott

SmartOwner is published (almost) daily by the team at DigitalTreehouse. Want AI consulting or automations for your business? Reply to this email.

Need Help With Your Digital Marketing?

DigitalTreehouse offers expert consulting, AI automation, and digital marketing services.

Contact Us Today

Related Articles