January 29, 2026
January 29, 2026
Mission North Account Director Lexy Cardiff reflects on years of hands-on work with AI: where it helps, where it doesn’t, and where human judgment still matters most.
I don’t use AI in my personal life.
I don’t use AI as a therapist, mentor, or coach. When I’m trying to figure out how to handle a human challenge, like preparing for a tough conversation, I’d rather talk to someone who’s been there, like a friend or manager. AI is a great tool, but it’s not a person.
Consistency is where AI adds incredible value.
AI is great at applying rules: brand language, naming conventions, formatting preferences, tone directives — the small things that matter to clients and are easy to miss when you’re moving fast. That’s where it earns its keep.
When you’re juggling multiple clients, industries, and documents, it’s unrealistic to hold every preference in your head. AI can. It doesn’t replace judgment; it supports it. It helps keep the work clean and consistent, especially under pressure.
I don’t start with AI.
I use AI to refine my work, not do my initial thinking. That first draft, where I’m figuring out if I know what I’m talking about or not, I need to own that process. If AI starts the work for me, I never know if I do.
When I’m stuck, AI helps get me unstuck.
Sometimes I can't think of the right way to say something. Or I need to condense a narrative into slide format. AI is genuinely helpful for that. It saves time on work that's time-consuming but not particularly value-additive.
It was one of the scariest moments of my life.
We were working on a survey report for a client. I had used AI to pull key findings from the raw data, trying to work faster and smarter. The draft was ready to go to the client. Then someone asked about a specific data point we had written up.
The data point didn't exist. The question hadn't even been asked in the survey. AI had hallucinated an entire finding, and I was about to submit it as a deliverable.
We caught it early enough to remove it before we were in too deep with the report and the narrative. But that sick feeling, realizing you almost put fabricated data in front of a client, that stays with you.
You need to gut-check everything AI gives you. AI has no gut.
When you're analyzing data, the biggest number isn't always the most interesting. If you know the client, the industry, and the context, you might see that the 20% figure is actually the story, even though 65% is bigger.
AI doesn't know that. It doesn't have the background. It doesn't have the instinct for what matters to media versus what matters to marketing. You have to bring that.
Good old-fashioned search is still incredibly important.
This morning, I wanted to know why Oklahoma City was hosting events for the Los Angeles 2028 Olympics. I could have asked AI, but instead I Googled it to get the actual article, not AI’s interpretation. I didn’t have to second-guess it.
You are responsible for your AI’s output.
If something goes wrong, the tool won’t take the blame. You will own it.
Interested in learning about how AI is used in media pitching? Let's talk.
