AI is transforming how creative work gets made — but most agencies still don’t know where to begin in using it.
Jules Love, founder of Spark AI, a consultancy that helps creative firms weave AI into their day-to-day work, says the real challenge isn’t technical — it’s psychological.
“Adopting AI in your team won’t happen by accident,” he told Business Insider. “You need to do it deliberately — make somebody accountable for it, and give them the space to be successful in that role.”
Drawing on his work with more than 60 agencies through Spark AI, he shares six ways leaders can future-proof their teams for the AI era.
1. Build an AI taskforce — not a tech committee
Love says most agencies won’t transform unless someone owns the AI integration work.
Rather than setting up vague “innovation groups,” he encourages leaders to assign responsibility and protect time for AI integration, even if that means pulling people slightly away from billable client work.
Agencies that succeed, he added, treat AI as a core business priority, not a side project that gets squeezed in when deadlines allow.
2. Make AI training role-specific, not generic
“It’s amazing how many agencies roll out ChatGPT or Gemini to their teams and don’t train anybody on it,” Love said.
He likens untrained teams to people staring at a giant box of Lego with thousands of bricks inside, but no picture on the front and no instructions. The picture on the box, he said, is role-specific training.
For him, training is what turns experimentation into practical capability.
3. Encourage play — and make it policy
Love said agencies struggle to innovate when teams are constantly under deadline pressure.
To make real progress with AI, leaders need to create structured time for experimentation — moments where people can test new workflows without the risk of missing a client delivery.
He pointed to companies like Lego, which regularly takes teams off-site to explore new ideas, and Canva, which paused normal work for a full week to rethink how departments could use AI.
“Fear kills innovation faster than bad tools,” Love said. “You have to give a little bit of room for failure.”
4. Replace fear with ownership
Love said a lack of openness around AI use is one of the clearest signs that something is wrong.
When employees feel the need to hide tools like ChatGPT from coworkers, it suggests AI is still seen as risky or illegitimate rather than useful.
To change that dynamic, he advises leaders to make AI use visible and normalized, especially by creating space for teams to share how they’re experimenting — including what hasn’t worked.
“That’s the No. 1 sign that there’s not a good culture around AI in your business.”
He also encouraged managers to push responsibility down the organization by giving individuals ownership over specific AI initiatives, such as maintaining a prompt library or developing custom tools, so adoption feels collaborative rather than imposed.
5. Shift your culture before your tools
Love said many creatives misunderstand how AI is meant to be used. Too often, teams treat it like a faster search engine — asking one-off questions and moving on — rather than as a collaborative assistant that improves with context and iteration.
He argues that real gains come when people learn to “brief “AI the way they would a colleague, giving it background, constraints, and feedback instead of quick prompts.
“It’s much better to think of it as briefing somebody else to do the job,” Love said.
6. Start small, but measure impact
Love said agencies risk undermining their own value if they focus only on how AI makes work faster.
As creative output accelerates, clinging to the billable hour can push firms toward commoditization rather than differentiation, he said.
Instead, he urged leaders to rethink pricing around outcomes, rather than speed, and to pilot AI in specific workflows so that teams can clearly measure what improves.
“If all we’re doing is doing more stuff faster, then we’re going to see a bit of a race to the bottom on fees,” Love said.
He believes that fixed-cost projects, which reward better results — not velocity — give agencies room to invest in learning and experimentation without eroding their margins.
A mindset shift
Love’s advice for 2026 is simple: “Stop thinking about what you can do today more quickly and what you can do tomorrow better.”
He believes the agencies that thrive will not be those with the biggest budgets or flashiest tech but those that help their people learn, lead, and experiment.
“Come 2027, you’re going to be looking pretty old-fashioned, pretty expensive, and pretty uninteresting as an agency if you’re not embracing this stuff and seeing what you can do with it,” he said.
