The Google Inside Adobe's Software Is the Same One Trying to Undercut It

The scare arrived on a Tuesday. On September 1, Jim Cramer posted on X that a new Google product "could be tough for Adobe....again," and the thing he was pointing at was Google Pics, an AI image-creation tool built on Google's Nano Banana model and aimed squarely at Adobe Express and Canva. The word "again" is doing real work. Google has been launching tools that undercut Adobe's pricing for over a year, and Adobe's stock now trades about 27% below where it started 2026.

Here is the part the headline leaves out. The same Google that is trying to take Adobe's business is also the company whose models run inside Adobe's most important software. Gemini, Veo, Imagen, and Nano Banana are licensed directly into Photoshop, Firefly, and Premiere Pro. Google is Adobe's rival outside the application and its supplier on the inside. That split is the whole story, and it tells you which side of it is actually threatened.

The dependency runs the other way

Map the chain from the bottom and the surprise appears. In the AI creative stack, the scarce asset is not the model. Adobe has bright-line alternatives across its apps: Gemini Omni Flash for fast ideation in Firefly, and Google's Veo, Runway, Luma, and Kling as selectable options in Premiere Pro's timeline generation tool, plus its own Firefly for anything that must be "commercially safe." When Adobe added that generative tool to Premiere Pro this month, it let editors generate shots directly in the timeline and choose which underlying model to use. That is the definition of a commoditized supplier base.

So the node that is hard to replace is not the AI. It is the assembled professional workflow Adobe controls — the timeline, the pipeline, the plugins, the muscle memory — and it is genuinely entrenched. Premiere Pro took an engineering Emmy in 2026, and around 85% of the films at that year's Sundance festival were cut on it. No retail app is displacing that. Google can sell a nearly-free picture generator to the consumer tier; it cannot walk into a Hollywood post house and rip out the editing system an entire team has used for a decade.

That is the inversion worth holding onto. In the popular telling, Google owns the frontier technology and Adobe is just reselling it. In the actual economics, Adobe owns the distribution and the switch-cost moat, and Google's models are interchangeable inputs it happens to license. Control of the workflow is the chokepoint; the model suppliers are fighting for shelf space inside it.

What the partnership actually costs

None of which makes Google harmless — just differently threatening. The risk is not that editors abandon Premiere. It is that Google both supplies the best model and gives a version of it away cheap, which caps how much Adobe can charge for its AI features. Nano Banana Pro is the marquee image-editing model Adobe now offers inside Firefly and Photoshop; Google Pics wraps the earlier Nano Banana in a free-ish Workspace product. Google is undercutting the price of the very class of image model it rents to Adobe, on the strongest tier Adobe has.

That is a real squeeze on monetization, and it shows up in the numbers. Adobe's Q2 fiscal 2026 quarter was a beat-and-raise: record revenue of $6.62 billion, up 13% year over year (11% in constant currency), non-GAAP earnings per share up 18%. Management lifted full-year targets. Yet the stock fell about 6% the next day, in part because Adobe said it would lean on a freemium model that pressures near-term revenue — the same free-tier creep Google is forcing on the whole industry.

The scale problem is the part worth dwelling on. Adobe said its "AI-first" annualized recurring revenue tripled year over year and passed $500 million. Impressive headline, small denominator: Adobe runs roughly $26 billion in trailing revenue, so its flagship AI metric is on the order of 2% of the business. The thesis that AI monetization will re-rate the company depends on that line scaling many times over, against a supplier that keeps pushing the price toward zero.

The market has already been told

The fears are not new, and the multiple reflects that. Adobe ended the summer at about $254, down 27% for the year and roughly a third from its 52-week high near $371, on about 14 times trailing earnings and four times sales. That is a deep discount for a company with 89% gross margins, a 40% free-cash-flow margin, and 12% revenue growth — a lot of the AI worry already in the price.

What changed the rating is a less gimmicky set of problems. In July, Morgan Stanley cut Adobe to Underweight and slashed its price target to $240 from $365, citing three "concurrent transitions": an expanding freemium strategy, leadership changes, and the AI shift in how consumers discover products. The leadership piece is real and unsettled. Shantanu Narayen announced in March he would step down after eighteen years as CEO, and on September 3 — two days after the Cramer post — Adobe named insider Anil Chakravarthy as his successor, joining the board effective November 16, 2026 and taking over as CEO on January 1, 2027. Anil runs Adobe's digital-experience and field organization, not the creative business under threat from Google.

So the honest two-sided read is this. The structural moat is intact: Adobe is the harder-to-replace half of its relationship with Google, protected by a professional workflow no consumer tool can dislodge. The economic squeeze is also real: the same model Google rents to Adobe is the one it gives away, which constrains pricing just as Adobe's AI business is still a rounding error on revenue. And the stock is neither cheap enough to ignore the risk nor expensive enough to punish it. For a beginner deciding whether this changes their Adobe and watch-list verdict, the deciding fact is which side they believe Google's models force Adobe toward over the next few years — a supplier whose prices only fall, or a partner the creative world cannot leave.