Somewhere around 2011, a marketing conference declared it “the year of mobile.” Boards were told that if they didn’t have an app by Christmas, they were finished. Websites were pronounced obsolete, replaced by something leaner, native, unrecognisable. It was, by all accounts, going to change everything.

The joke, told fondly, is that the year of mobile ran for what felt like 4-5 years. Mobile was genuinely transformative, but it meant something quieter than the keynote suggested: another screen size to design for, another good tool in a kit that already contained several. The website survived, it just started working harder, in more places, for more people. It became part of the whole rather than mobile, a replacement for it.

I think we’re living through the same story again, with a different noun in the headline. AI is not overstated, if anything it’s under-described. Creative production has sped up to near-instant iteration, data processing that used to take a fortnight now takes an afternoon, and it’s embedded as a feature inside almost every product you already use, from your CRM to your bank’s chat window. Worth a note of caution, however that the entire stack we commonly use rests on a handful of companies whose finances are coming into question in the mainstream media. Watch this space.

AI is huge, and I’m not taking away from it, my issue is with the shape the hype is taking. Every important new technology gets discussed in its early years as though it will replace everything before it, and it never quite does. It gets absorbed, messily, into how work already gets done. The marketers who did best out of the year of mobile, kept delivering on objectives that didn’t shift with the hype. I suspect AI will do the same, it’s just another tool in the utility belt and not a single source of research or buying. Treating it as the only tool is where we’re already seeing issues, from SEO performance dips to homogenised creative adding to the sea of sameness. AI can deliver an output that meets your prompt but outside of the surreal, it still takes a human to deliver something truly new and with emotion that cuts through.

The end-user side is where financial services earns its own chapter, because this is a category built on decisions people don’t make lightly. AI is very good at narrowing a hundred options down to three. But watch what happens at a genuinely serious financial decision, a pension provider, an inheritance, a mortgage. People still want to talk to a person, land on a website with a phone number at the top, and check the decision a second time before committing. AI gets you there faster and better informed, but much like a trip to the doctor’s, you often want a second opinion with the expert before moving forward.

This will shift as agents become more capable. But there’s a meaningful gap between trusting an agent with a decision you can reverse by Friday and one you can’t reverse at all. Nobody’s handing their pension consolidation to an agent with the ease of a restaurant booking, not this side of 2030 anyway. Financial services brands that understand this have a genuine opportunity to be the calm, human, findable thing at the end of an AI-assisted research journey. Ideally, this is within their own website environments or apps, not only from a trust building perspective but also control the fine line in terms of targeted support.

So, practically, bring AI into the everyday. Speed up the mechanical parts, creative variants, first-pass analysis, so more time goes toward judgement and trust-building, rather than treating AI as a wave about to flatten the industry. However, scrutinise and maintain the human layer as we’ve all seen some of the questionable outputs still coming out from LLMs. We shouldn’t rely on it fully, even if some clients and prospective customers might. Let’s create the space where they can get support, on our own turf, in our own way, as part of the brands experience. 

The year of AI, like mobile, will run for a while yet. Bandwagons come and go, and they all look unmissable from the inside. There’s real opportunity in leaning into AI but the real risk, is in leaning on it.