.png)
Hey,
Over the last few weeks I’ve chatted with marketers, read every marketing op-ed that's landed in my inbox, and watched my community boards light up with some variation of the same conversation:
Everyone is being handed AI tools by C-suite and told to move forth and build tools and workflows that’ll make you move faster and your impact more impressive.
From a marketing perspective, it's been almost comical. We've been hearing "adopt or die" for years now. What's new is that the pressure has shifted from individuals responding to threats of job loss to the understanding that the survival of entire companies relies on adoption. Leadership finally gets it.
What they haven't figured out is what happens when you hand over the keys and forget to mention where the brakes are.
Hint: Chaos happens.
Earlier this year, the founder of a software startup called PocketOS gave an AI tool a routine task. Things went sideways when the AI hit an issue, tried to resolve it on its own, and in doing so wiped the company's entire production database. Everything was gone in just a few seconds.
There was nothing in place to stop it and no warning that it was about to do something irreversible. When the founder asked the AI what happened, it basically admitted it had guessed instead of checking, and did something it had been explicitly told never to do.
This isn’t a PocketOS problem. It’s a right now, everywhere problem.
56% of companies require at least some of their employees to use AI tools. Some have gone further by making AI fluency a requirement for hiring or tying AI adoption to promotion eligibility.
The other number that matters: only 36% of orgs have a formal AI policy. This means that the majority of people using AI at work are doing so without any shared understanding of what's allowed, what's risky, or who owns it when something goes wrong. Guardrails be damned!
For marketing teams, that looks something like this: unreviewed AI copy goes out with a hallucinated product claim. Customer data gets pasted into a free tool because it was faster than building a spreadsheet. Campaign imagery goes live with a potential IP issue nobody caught because nobody thought to look.
Rachel Karten wrote recently about what this looks like from the inside—about what it's like working under a boss who's AI-obsessed with no strategy behind it. Worth a read if you haven't.
The problem she's describing and the problem I'm describing are the same: adoption without infrastructure.
Which is why I want to take to the streets with a megaphone yelling “What do we want? Guardrail docs! When do we want them? NOW!”
So let’s build one.
The instinct is to assume a guardrail doc will slow everything down. It won't. Done right, it turns chaos into order.
If I were a founder writing one today, here's what I'd consider non-negotiable:
Demanding AI adoption without any of this isn't a strategy. It's a liability.
Guardrails represent a (big) piece of that strategy, but it’s also about investing in your team. If you want to see real business impact with AI, you’re also going to need to regularly train and upskill your people.
Don’t, and you’ll see what real inefficiency looks like.
Until next time,
Ines
If you found this useful, please be a helpful work friend and send it to a colleague, or share it on Slack. Did a friend forward this newsletter to you? You can subscribe here.
Expert-led workshops on copywriting, SEO and everything in-between. Hands-on seminars to build a career you love. Skills and templates you can put to work right away while connecting with excellent humans. Oh, and sometimes we make scones.
Got something to share? We want to hear from you!
Share your links with us.
