September 25, 2026 · Daily Radar
"Using Photoshop from a chat box makes designers unnecessary." Since last night I've been reading different versions of that sentence. I disagree. But my objection isn't meant to downplay the news, because what's really changing isn't the designer's job; it's the tedious tasks a small brand has to get done every day without going to a designer.
On September 24, Adobe announced two things. Photoshop, Lightroom, Adobe Express and Firefly tools came to Google Gemini. The same day, Acrobat was added for the first time to the Adobe connector in Anthropic's Claude, pushing the connector's tool count past 80 [A]. That means fixing the lighting on a product photo, adapting an image to three different social media sizes, or deleting a page from a proposal PDF and blacking out personal data can now be done by typing, without opening an app. On paper, at least.
Quick answer: On September 24, 2026, Adobe opened Photoshop, Lightroom, Express and Firefly tools to all Gemini plans, and added Acrobat to its Claude connector. It matters most to e-commerce and service businesses without a design team: repetitive tasks like product photo fixes, resizing and PDF editing can be done through chat. Try it first on a task that involves no customer data, then set a written rule about which files may be uploaded to these tools.
According to the announcement on Adobe's official blog, "Adobe in Gemini" began rolling out globally on September 24 and is available on all Gemini plans [A]. To turn it on, you go to the Personal Intelligence section in Gemini, pick Adobe from the connected apps list, and sign in with your Adobe account. If you don't have one, you can create a free account.
The use cases listed in the announcement are: making lighting and color consistent across a series of product photos, adapting a single design to different social media formats, and customizing Express's ready-made templates for a campaign [A]. Coverage from PetaPixel and 9to5Mac confirms the same scope [B].
The Claude side is a bit different. Adobe already had a connector there; this time it expanded. Acrobat arrived for the first time: deleting, merging and splitting pages, blacking out sensitive information (redaction), and converting PDFs to other formats. There's also an interactive PDF editor; you can reorder pages, fix text, and add comments and highlights. The connector works in Claude and Claude Code, on desktop, mobile and web. You can start as a guest, but full features require an Adobe account [A].
For the Gemini side, Adobe calls this "the first step in our journey" [A]. That sentence matters. Today there are four apps in Gemini and nine in Claude. When that gap will close isn't clear, and I'm not going to guess.
The numbers part is short. Adobe writes that Gemini reaches more than a billion people a month [A]; that's a company figure based on Google's own claim, not an independent measurement. The Claude connector covers more than 80 tools and nine apps [A].
I think the real significance isn't in the numbers but in where the work happens. Until now, visual work at a small business went like this: someone takes the photo, someone else (often an agency or a freelance designer) fixes it, cuts it into sizes and sends it back. Two days pass in between. Now the repetitive steps in the middle of that chain are moving into a chat window that's already open every day anyway.
That doesn't mean the designer is done. Brand consistency, meaning an image looking like "us," doesn't come from a sentence typed into a chat box. You have to explain your color codes, your typeface and where your logo goes every time, and the announcement doesn't say whether the tool will remember that in the next session.
There are open questions too. PPC Land's coverage lists them one by one [C]: how long uploaded files are kept and whether they're used for model training, whether Google Workspace accounts are covered, restrictions in the European Economic Area, and whether Firefly usage is deducted from the generative credits in an Adobe subscription. Adobe's own materials also contradict each other on whether an Adobe account is required or optional [C]. The footnote of the announcement says "Compatibility and availability varies," and that's it.
How well it works with prompts in Turkish isn't clear yet; there aren't enough real-world trials.
| Task | Yesterday's way | After September 24 | Watch out |
|---|---|---|---|
| Fixing light/color on a product photo | Someone who knows Photoshop | By typing in Gemini (Photoshop, Lightroom) | Check series consistency by eye |
| Adapting one image to 3 social sizes | A designer or manual cropping | With Express in Gemini or Claude | Text and logo may get cut off |
| Campaign template | By hand in Express/Canva | Customizing the template through chat | Give the brand colors every time |
| Deleting or merging PDF pages | An Acrobat license | Acrobat in the Claude connector | Features are limited in guest mode |
| Redacting personal data in a PDF | Acrobat Pro | Redaction in the Claude connector | Uploading the document is a separate KVKK decision |
If you'd rather hook these tools into a recurring task than try them one by one, I explained how to build your first workflow in the n8n and AI automation guide.
Pick a single product or service image that contains no customer data. Turn on the Adobe connection in Gemini and have it adapt the same image to both square and vertical story formats. Put the result side by side with the version a designer or you made by hand, and answer one question: would I publish it as is? If the answer is "yes," you can hand that task over to chat. If it's "no," note what's missing; you'll have learned what the tool can't do.
If you're wondering which of your repetitive tasks can really be automated, we explain how we map them on the AI automations page.
| News | Why people are talking about it | Who it concerns |
|---|---|---|
| Google's September 2026 spam update began (September 24) | Google wrote that the rollout could take "up to two weeks"; this year's previous three spam updates finished in 1-3 days. What it targets wasn't disclosed. | Every site with organic traffic. Don't panic over daily ranking swings for a week. SEO management |
| Search Console got a "multimodal" search type filter (September 24) | Traffic from Lens, Circle to Search and image-based searches shows up separately for the first time; query text isn't shown for this traffic. | Image-heavy e-commerce and anyone measuring AI visibility. GEO agency |
| "Holistic" and "Growth" budget modes on the Google Ads recommendations page (September 24) | Holistic takes budget from weak campaigns and moves it to strong ones; Growth suggests extra budget without cutting anything. The second means more spending. | Advertisers who apply recommendations with one click. Google Ads |
| An AI test in Google Business Profile that pulls opening hours from plain text (September 24) | When you write "weekdays 9-6, half day Saturday," Google fills in the hours itself. It's a test for now. | Local businesses; wrong hours mean customers lost in map searches. Digital marketing |
| Acrobat came to the Claude connector (September 24) | The other half of the headline: PDF editing and redaction can be done through chat. | Service businesses working with proposal and contract files. Enterprise AI consulting |
According to Adobe, the feature is available on all Gemini plans, and anyone without an Adobe account can create a free one [A]. The announcement doesn't say whether use of generative tools like Firefly is deducted from Adobe's credit system; until that's clear, don't assume "free and unlimited."
The announcement says "global," and there's no wording that excludes Turkey [A]. But the rollout started on September 24, so if it isn't showing in your account yet, wait a few days. Adobe's footnote also notes separately that availability varies.
Redaction requires uploading the document to the AI service. How long files are kept and whether they're used for training wasn't disclosed in this announcement [C]. Documents containing personal data first need an internal company rule and a legal opinion.
Let's work out together which of your image and document tasks are a fit for automation: AI automations · Production and video · For another change in AI tools last month: The Sora 2 API is shutting down
On October 8, 2026, Google introduced a Gemini work agent that runs with its own email address and audit log; pricing and a general availability date were not announced.
From a workflow that writes form responses to a spreadsheet to appointment reminders on WhatsApp, here are the ten workflows businesses most often build with n8n, each with its trigger, its steps and what to watch out for.
I explain which tasks are worth automating, how n8n works and whether it is really free. You'll also find how it compares with Zapier and Make, the rules of WhatsApp automation, and a checklist to run before you build your first workflow.
We use cookies to improve the site and measure advertising performance. Analytics and advertising cookies run only if you accept them. Details are in our Privacy Policy and Data Protection Notice.
