“Innovators make the previously impossible possible.”
—Max McKeown, Author
The AI Breakdown
The Cybercab's First Day Came With a Federal Audit
Tesla started charging fares for Cybercab rides in Austin, TX on September 3rd.
Federal regulators opened an investigation the same day.
The National Highway Traffic Safety Administration's focus is whether Tesla followed the right process when it declared the vehicle street-legal. The Cybercab has no steering wheel, no pedals, and no mirrors. Federal safety standards were written for cars that have all three.
Under U.S. rules, automakers certify their own compliance and sell without advance NHTSA approval—essentially grading their own homework before putting it on the road. Tesla decided the Cybercab meets every applicable standard and started taking passengers. NHTSA is now auditing that decision.
Tesla / The Autopian
The contrast with Amazon's Zoox is worth noting. Zoox built a similar vehicle and applied for a formal federal exemption before charging a single passenger for a ride. That exemption just came through, allowing Zoox to launch a paid robotaxi service in Las Vegas. Tesla looked at the same regulatory landscape and decided it didn't need to go through that process at all.
One company asked permission. The other decided it didn't need to. Now NHTSA is weighing in on that decision—and it can require design changes, demand a formal exemption, or even pursue enforcement action.
Tesla wants to scale Cybercab to millions of vehicles. How fast that happens now depends on how NHTSA evaluates the paperwork.
Prompt of the Week
The Cybercab story is a good reminder that when new technology moves faster than the rules written to govern it, the companies watching from the sidelines have a decision to make: wait for clarity, or start preparing now.
Most businesses have at least one AI tool, vendor, or workflow that falls into a similar gray zone. The question is whether you know where yours are.
Act as a regulatory and operational risk advisor.
I am going to describe an AI tool or automated workflow currently in use at my organization:
[DESCRIBE THE TOOL OR WORKFLOW]
It is used by:
[SALES / MARKETING / OPERATIONS / FINANCE / LEGAL / HR / OTHER]
It touches or influences:
[CUSTOMER DATA / FINANCIAL DECISIONS / HIRING / COMMUNICATIONS / COMPLIANCE / REPORTING / OTHER]
Review this tool or workflow and identify:
1. Where the rules or regulations governing this use case are still unclear or evolving
2. Where my organization may be self-certifying compliance without formal validation
3. What a regulator, auditor, or legal team might flag if they reviewed this tomorrow
4. What steps I should take now to reduce exposure before the rules catch up
5. A plain-English summary I could share with leadership
Be specific. Use examples from my industry where possible. Flag anything that looks like a shortcut I may not have noticed.
Fresh Finds for Auto Pros
BDC & Customer Communications: Alma
Most voice tools run on general-purpose AI that wasn't designed for live phone calls. Alma was—its parent company trained it on more than 10M real conversations. For a store fielding calls about appointments, loaner status, and repair updates all day, a bot that actually keeps up with the conversation is worth paying attention to.
Marketing & Advertising: PureCars
A customer data and marketing platform built for dealerships. It connects first-party data, AI, and media buying in one place. So, instead of running campaigns based on broad audiences, you're targeting people who have already shown real buying signals on your own website and in your CRM.
Content Creation: Creatify
An AI video ad generator. Drop in a URL or product description and it produces a short video—script, voiceover, avatar, stock footage. Useful for testing a lot of creative variations before involving your production team.
Just For Fun: Infinite Slop
A live AI TV channel where the audience decides what happens next by typing prompts into a chat. Clowns eating Play-Doh at a fancy restaurant. A giant singing banana winning America’s Got Talent. A high-energy infomercial-style video of an alien selling intergalactic cell service. Absolutely unhinged stuff. Go waste five minutes.
Hear from the Experts
AI can make messy data look usable, but that gets a lot more dangerous when you expect AI to actually run dealership workflows.
AJ McGowan, VP of Research and Development at Reynolds and Reynolds, explains what changes when dealership AI moves beyond chatbots and analysis into truly agentic systems. His argument cuts against an increasingly common assumption: just because AI can figure out relationships between disconnected data doesn’t mean you should make it solve that problem every time.
Bits and Bytes
A Microsoft engineer says “typing code is absolutely over” as Copilot moves deeper into the full developer workflow. ⌨️
Another OpenAI rogue agent incident has surfaced, and the bigger issue may be who gets to decide what counts as a full investigation. 🕵️♀️
DeepSeek is raising $7.4B at a $74B valuation—up from $50B just a few months ago. 💰
Parting Pixels
Thanks for reading, Friend! This is the era where one person with the right tools can do what used to take a whole team. Use them to your advantage.
