Perplexity just struck a multi-year deal with Getty Images to license and properly attribute photos inside its AI-powered search tools. This is the first real sign that AI's wild west era might be ending.

Until now, most AI tools have blurred the line between inspiration and infringement. Perplexity itself has been sued roughly a bajillion times.

Adobe has tried (maybe not hard enough) to keep its hands clean with Adobe Stock as a training source for Firefly, but the results, and adoption, have been mixed. Shutterstock has long had a similar partnership with OpenAI, licensing its image library for model training and compensating contributors when their work is used.

These moves point to a larger reckoning.

Every generative platform will eventually have to pay the creative class (think the measly revenue generated by most artists on streaming platforms like Spotify), not just as a moral gesture, but as a market correction and future cost of doing business. The Getty-Perplexity deal accelerates that timeline.

For marketers, it exposes a gray zone that’s thankfully beginning to see some clarity. Using AI-generated visuals for campaigns, ads, or web creative might still feel cost-effective, more relevant or precise and even higher polish, but without a clear licensing trail, it’s becoming a brand-risk decision, not just a creative one.

Attribution is shifting from a “nice-to-have” to a compliance issue.

For CMOs, this changes the calculus around AI-driven creative. What started as an efficiency play with faster ideation, cheaper visuals, endless variations, now intersects with brand risk, legal exposure, and trust.

Dozie Anyaegbunam thinks no.

And he’s got a number of good points to make.

In his piece “The Marketing Funnel Doesn’t Exist,” Anyaegbunam argues that the familiar awareness / consideration / decision model comforts marketers more than it reflects reality.

Real buyers don’t glide through stages, they move in bursts triggered by moments, not nurtures.

Buyers buy when something shifts, not because your next email finally convinced them or your tactics ebook gave them new perspective.

This shift in thinking reframes the marketer’s job. Funnels focus on motion. Marketing should focus on memory. Top-of-mind brands aren’t the ones with the most sophisticated nurture flows, they’re the ones already sitting in the buyer’s head when the trigger hits.

Being easy to recall, easy to trust, and easy to find at the moment of need should be our priority as marketers. This perspective necessarily places more emphasis on the increasing importance of brand in marketing, ensuring what you say is as consistently engaging as it is distinctive.

Buyers don’t move through funnels, they move through moments.

Brand isn’t the garnish on demand generation. It’s the memory system that makes demand generation work at all.

We help B2B teams cut through the noise with sharper narrative, clearer positioning, and content systems built for distinction, not duplication.

Nearly three decades after Clippy’s unwelcome interruptions came to an end, Microsoft is once again trying to give AI a face, or at least, a feeling. Enter Mico, the expressive orb now living inside Copilot’s voice mode. It reacts to tone, remembers context, and even teaches through Socratic dialogue.

This is smart brand work disguised as product design. Just like crypto before it, AI desperately needs its dystopian edge softened. The temptation to lead with scale, power, and processing speed, the “cold” side of tech, is a trap too many B2B brands eagerly fall into.

Microsoft is taking the opposite route: embedding warmth and personality directly into the customer experience.

And it’s not without self-awareness, which in itself is quite charming. As many have pointed out, the Microsoft is embracing its own history, nodding to Clippy with a wink, even teasing an Easter egg for nostalgic fans (and maybe a knowing laugh for anyone who remembers this classic “Wait Wait... Don’t Tell Me” rant about Clippy’s overzealous cheerfulness).

Whether it's successful or not remains to be seen, but Mico is an example of Microsoft getting the process right: leading with an outcome in mind, rather than with features and benefits.

When you build your brand and messaging around features and benefits you get what most competitors in the AI space feel like: an audition for Black Mirror (ChatGPT's voice mode, anyone?).

According to McKinsey, AI agents could reduce software seat counts by as much as 70%, effectively ending the per-user pricing model that’s defined SaaS for two decades.

AI is changing who and what your customers are. Software will increasingly be bought, used, and optimized by agents, not people. That means fewer human “users,” fewer logins to measure engagement by, and a marketing ecosystem where usage is the new metric of success.

This shift will also force B2B brands to sell outcomes, not access. Vendors who once marketed “seats” will now have to prove ROI in real time, showing the actual business impact of each token, workflow, or agent decision.

Some thoughts on how to prepare for the shift:

This is a new kind of buyer requiring a new kind of pricing model, a new kind of proof, and a new kind of marketing built around value creation at machine speed.

Every metric matters until the moment someone has to feel something. That’s where brand begins, and where most automation ends.

—Jay