Question: My boss has become obsessed with AI since attending a conference. Now every time we bring up a marketing challenge, he says, “Just run it through AI!” He thinks it can instantly analyze customer behavior, predict campaign success, rewrite our entire website, and edit videos — basically magic. How do I set realistic expectations without sounding resistant or unenthusiastic? 

Answer: Before we dig into the challenge, let’s start with the positives — because there are some important ones. 

First, your boss is excited about testing and implementing new technologies. That, in and of itself, is a huge win. Many leaders are reluctant to try new tools or techniques. They worry about potential problems, prefer familiar processes and choose to sit on the sidelines while other organizations take the risks and reap the rewards. 

Second, your boss didn’t just delegate “figure out AI” to someone else. He chose to attend an AI-focused conference and learn firsthand. That matters. Instead of assigning a vague mandate like “improve our workflow with AI,” he invested his own time so he could participate in the conversation. How many leaders do you know who would do that? 

The friction you’re feeling comes from a very common place: the tension between enthusiasm and expectations. 

AI has been positioned — especially in marketing — as a wonder technology that can do no wrong and only make life easier. Conferences tend to emphasize potential rather than process. Attendees see best-case scenarios, polished demos and impressive case studies, not the messy realities of implementation. It’s no surprise your boss came back wearing rose-colored glasses. Anyone would. 

Your perspective, however, is different — and equally valid. 

You understand that AI is a tool. A powerful, disruptive tool with the potential to reshape how marketers create and execute campaigns, but a tool nonetheless. Your focus is practical: where AI fits into the day-to-day workflow, what inputs it requires, and what constraints still exist. 

Your boss is looking at the 50,000-foot view — the company, the industry and the long-term potential of AI. You’re on the ground, dealing with data quality, context and execution. Neither of you is wrong. You’re simply solving different problems from different vantage points. 

So, how do you communicate that without sounding negative or resistant? 

Start by reframing the conversation rather than pushing back against it. When a leader says, “Just run it through AI,” they’re usually responding to a desired outcome — speed, efficiency, insight — not laying out a detailed plan. Instead of challenging the idea, focus on clarifying the goal. Once you’re aligned on what problem you’re trying to solve, AI becomes one possible path forward, not a magic shortcut. 

It also helps to reposition AI as an accelerator, not an answer. AI can help generate a first draft, surface ideas or summarize information quickly — but it still needs direction. It doesn’t understand your audience, your brand or your business constraints unless you teach it. Framing AI as something that supports human judgment, rather than replaces it, keeps expectations grounded without sounding dismissive. 

In the moment, language matters. Simple, collaborative phrasing can reset expectations while preserving momentum. Statements like, “We can absolutely use AI here — what part do you want it to help with?” or “Let’s use AI to speed up the first pass, then refine it,” keep the conversation constructive and forward-moving. 

Finally, when expectations start to drift, suggest small experiments instead of big promises. Piloting AI on a single task or campaign — with clear goals and defined human checkpoints — turns abstract excitement into something concrete. It also creates shared learning, which is far more productive than debating what AI can or can’t do in theory. 

Being realistic about AI doesn’t make you anti-AI. It makes you effective.