In January 2026, Google unveiled new technology that streamlines the consumer search-to-buy experience. Why? Because if Google controls the distribution layer, Google controls lending and purchasing, too. Credit unions better take note.

Just a few weeks into the year, Google didn’t just validate those pointing to Agentic AI as a key trend for 2026, but upped the stakes significantly. In one fell swoop, Google’s Universal Commerce Protocol (UCP) elevated their AI from being conversational to functional, opening up a new front credit unions may be ill-equipped to fight on.

Whether this made it into your news feed or not, make no mistake: this was a Rubicon moment.

So long SEM?

To be honest, the UCP model is simple: Consumer A uses Google to search for Product B, Gemini (Google’s prominent AI interface) offers recommendations and in one click the item is purchased. On the face of it this is nothing revolutionary, right? But let’s look a little closer.

The rub here is that UCP places the entire shopping experience inside of the Gemini layer, and what Google suggests, people buy.

That power of influence is why institutions invest billions (trillions?) in SEM, after all. So if Google’s willing to risk its SEM revenue, what’s the prize?

Distribution is up for grabs

In this first iteration, relatively straightforward retail seems to be the testing ground for UCP. It’s built for “I want a new black jacket,” rather than “Please lower my mortgage” because tangible goods have structured inventory, straightforward pricing, and checkout is typically a formality. The potential for scale, however, is clear.

Auto is a near-term example of where this could infringe on traditional credit union territory. Vehicle inventory is already machine-readable, monthly payment calculations are automated, and prequalification APIs exist, so it’s no great leap for a consumer to type, “Find me a 2022 RAV4 under $25,000 for less than $400 a month,” and have the solution plopped in their lap.

Mortgages would be more complex but Google isn’t trying to replace underwriting or title work. It doesn’t have to. That’s grunt work. No, the Google Gods want domain over the distribution layer. That’s where the money is.

Now, whether they know it or not, anyone Googling “Best way to finance a car” is not peripherally browsing so much as actively delegating. Gemini (or ChatGPT, or, or, or…) will search, source and recommend products, and, critically, provide a step-by-step guide on how to move forward. If a credit union’s solutions aren’t in that mix, they’re not playing the game.

So who pays the piper?

Perhaps the wildest thing about this is that Google was actually a little late to the game. OpenAI launched an embedded shopping experience in late 2025 and that extra calendar quarter gave them enough time to figure out financing. In January they announced that transactions completed inside ChatGPT would carry a 4% platform fee (interesting note: ask ChatGPT about this; it will insist it’s a “potential” fee).

Likely this is an indication of how this experience will monetize itself: if you own the bakery, you get a slice of the pie. In retail, that will be a transaction fee layered on top of payment processing. For financial services? We wait and see, but while the specifics may differ, I’ll bet you a Google-sized fee the principle remains the same.

Immediate and long-term scale shake out the same economically

That Google-sized fee is a bet I’m willing to make because when this model scales, it’s going to cost credit unions one way or another.

See, Google describes UCP as an “open” standard. OpenAI class their AGC (Agentic Commerce Protocol) as “open”, too. That means there’s nothing proprietary about the zeroes and ones there; the plumbing’s all in-situ, just hook yourself up! But Google didn’t hook up to AGC in September 2025, did it? It launched UCP instead.

If that pattern persists, distribution becomes completely fragmented. Google, OpenAI, Apple Intelligence, Microsoft, X etc. they all control their own fiefdoms, and can write their own rules—and cost structures.

Roll the tape forward and that means there are fees every way you cut it. You either pay to play on each individual platform, or a dominant force can charge because of its market share and distribution scale. That game’s called They Win.

Reassess product structure, be honest about value

Even if financial products aren’t flowing directly through these pipes today, the direction of travel is obvious, and credit unions need to start adjusting their product roadmap and infrastructure accordingly.

To compete, credit unions need more than a decent SEO strategy, slightly evolved from the keyword stuffing of yester-year. Showing up in the distribution layer will demand a more holistic approach to planning, building and rolling out financial products; and doing so with unabashed clarity about their value, quirks, when and where they work, when and where they fall down, etc.

Honestly, taking that approach is something credit unions should jump at. We should be this honest. We should offer nuanced, specific products for nuanced, specific markets. We should be absolutely transparent about fees, rates, risks.

This can stand the system apart from traditional banking counterparts who make money in the mirk, and could also be a great leveler within the credit union movement.

Ultimately, the question is the same as it always is: there’s a chance here, who’s willing to take it?

Looking to stay relevant? Reach out to discuss how we can help.