The penultimate month of the year brought some interesting AI in marketing updates. Google dropped Nano Banana Pro, built on their also all-new Gemini 3 — and some say it’s the best AI image generator we’ve seen yet. TikTok is giving users more control over their feeds when it comes to AI content. Recent surveys say returns on AI investments will come, but slower than what the market currently indicates.
Here’s what happened in November.
Subscribe to the ai marketer
Weekly updates on all things ai in marketing.
Google Released & Quickly Implemented Gemini 3 & Nano Banana Pro
On November 18th, 2025, Google announced and released Gemini 3, the “Strongest model in the world for multimodality and reasoning,” according to their reveal video. I won’t bore you with the benchmarks, but if you’re curious, check out Google’s Gemini 3 blog linked above.
Just a couple of days later, on November 20th, Google announced Nano Banana Pro, its new image generator that boasts higher quality, more in-depth visuals that go beyond just “beautiful images,” and helps users create “more helpful content,” such as infographics and diagrams.
One Reddit user says, “Nano Banana Pro is the first generative AI model where I incorporate its results into my job. I give it simple SketchUp models and 2D architectural drawings, and the results are usable. Realistic and accurate renders in seconds.”
Within this cascade of updates, Google also introduced a new SynthID invisible digital watermark technology. Users can upload an image to Gemini 3 and ask it if Google AI generated it. This is interesting, and I hope it sparks a trend across other platforms and AI models. As AI-generated images continue to get better, having that transparency will be more important.
Slim marketing teams and small businesses without graphic design resources might finally have a solid AI option to lean on for assistance. I haven’t tried it myself, but I’ve seen some side-by-side comparisons with other popular image generators, and Nano Banana Pro looks pretty impressive. Its realistic images (contingent on the prompt) don’t seem to have that fake cartoony look like other tools — they look more imperfect in a true-to-life kind of way, not in a bizarre AI way.
Nano Banana Pro is available now across most of Google’s products that support Gemini, including in Google Ads.
TikTok Gives Users Control Over AI-Generated Content
I’m glad we’re starting to see platforms and organizations give more control to users when it comes to AI content. Google’s SynthID watermark technology enables users to explore the origin of uploaded images (see above), and TikTok now enables users to control how much AI-generated content (AIGC) they see in their feeds.
Within the platform’s Manage Topics tool (Settings > Content Preferences > Manage Topics), users can adjust a slider to choose how much AI content they see while using the app. This move is in line with TikTok’s stance on AIGC, as they already require users to label realistic AI content, including images, audio and video, as part of its Community Guidelines.
This is a win for users, and I hope more platforms follow suit.
It’s unclear whether changing this setting affects sponsored content and ads, but if you’re a marketer who advertises on TikTok using AI-generated content, you might see impressions fall as users tweak their settings.
AI Investment Far Exceeds Revenue
A recent analysis from The Economist suggests that, despite massive capital flowing into AI infrastructure and tools, “normal business” adoption is lagging. The expectation that AI would rapidly transform everyday business operations — boosting productivity, profits and efficiency — is not materializing at the speed many had hoped.
While many marketers say they’re using AI at work, few companies have yet to effectively operationalize AI meaningfully. This is kind of a nice reminder that good things seldom happen overnight, and that improvement takes time.
Businesses and marketers shouldn’t expect instant productivity jumps, but rather gradual improvements that compound through small, well-integrated use cases.
A Quick Update on Anthropic & the Automated Cyberattack
In another blog post, I covered what’s been happening with Anthropic and what they called the first documented case of an cyberattack. If you want to get caught up on what happened there, that blog can bring you up to speed.
Fast forward to today, and Anthropic’s CEO, Dario Amodei, will appear in court on December 17th, 2025, to testify. I’m curious how that will all unfold, and what new rules or regulations might come of it to help protect people and businesses against automated cyberattacks.
The Countdown to 2026 Begins
As we wind down toward the end of the year, AI companies and users haven’t stopped innovating and experimenting with a wide range of use cases and capabilities. Google alone released many new products and features this month — enough to keep curious marketers busy for a while.
Google and TikTok’s moves to be more transparent about content origins and give more control to users, respectively, may instigate further-reaching trends that other platforms and tools adopt that give more power to the people. Only time will tell — and we’ll be on the lookout for these and other AI developments that impact marketers.
