Adobe can scan user-generated content stored in its cloud, and users aren't happy

Alfonso Maruccia

Posts: 1,127   +337
Staff
A hot potato: Photoshop maker Adobe changed its terms of service earlier this year, and creatives have since discovered that the company now states its right to freely scan user-generated content. However, the new terms are not significantly different from what they were before.

Video game concept creator Sam Santala recently discovered that Adobe has imposed new "Terms of Use" for its tools since February 2024, stating that the company has "full access" to content, including NDA-protected documents. Santala expressed his surprise about the new terms, threatening to cancel his subscription and switch to less intrusive Photoshop alternatives.

Adobe's General Terms of Use were last updated on February 17, with changes to a couple of paragraphs explaining how the company can (and likely will) access content and how that content is defined. Adobe may use manual or automated methods to scan content using machine learning algorithms, the sections state, but only in limited ways as permitted by law.

Scanned content includes text, audio or video files, documents, images, and essentially everything else that's been uploaded, imported, or embedded through the cloud-based software services sold by Adobe. The company states that it uses the results from this scanning practice to improve its services. Additionally, it could use different technologies and methods – including manual reviews – to identify CSAM material or spam activities..

Adobe's terms of service feel like a fairly standard business practice for a cloud-focused corporation, and the February update didn't bring much change to TOS documentation that has been enforced for years. If anything, Adobe has been scanning user and creator content for a long time, with a limited opt-out option for business-level accounts.

Adobe tried to reassure subscribers about its scanning practice, with the company's Chief Strategy Officer Scott Belsky stating that no generative AI models are trained on customer content. "Tight" security measures protect content from any form of access, Belsky said, and Adobe "will never assume" ownership of anything saved on its cloud servers.

The company also published a clarification post related to its TOS, trying to clear up any confusion about the updated terms and how user-generated content is managed on the cloud platform. There are probably circumstances where the terms allow Adobe "some degree of access," Belsky said, but the practice is seemingly designed to work in users' favor.

Permalink to story:

 
You left out the part that says Adobe can freely use and sub-license what they scan. Doesn't sound like internal use to me. Did the original license contain this line as well? I'd find that hard to believe.
The great part about using massively over broad licensing terms is that even if they don't use the data externally, they don't have to change the terms to start doing so.
 
They could always reference what you created so technically they don’t own it but that won’t stop them from copying/referencing your work I imagine and incorporate just some minor differences.
 
A lot of things were happening even before the internet such as data collection and others. Only difference is the internet made things more widely known and easier for companies to do things.
 
I will NEVER rent software, but I do have to use Adobe Audition at work. My employer pays the obscene charges involved. The first thing I did after setting Audition up was to change the settings so that everything I do is stored locally, and I then disabled Creative Cloud. Just store everything locally... Duh...
 
This is why I have NEVER used Adobe's cloud sh*t. I routinely close all of Adobe cr*apware every time I reboot or use an Adobe product. I resent that I need to do that. I can get by, just fine, with an occasional manual update check and nothing more. but no -
Adobe wants to know *everything* going on my 'puter. Sadly, I need Adobe stuff. Nothing else will do the jobs properly.
 
Gee, store something on a cloud server, and "shocked" whomever owns the server might scan it?
It's at the point now where they are planning to record everything you do in app as well. So even if you stored it locally they will see it get made. All in the name of training GAI
 
Back