The Google Is Very Unhappy: People Are Copying Its Work. Illegal! Unfair! Terrible!
February 18, 2026
Another dinobaby post. No AI unless it is an image. This dinobaby is not Grandma Moses, just Grandpa Arnold.
Up front I want to point out that I don’t know if the story is spot on. I am assuming that it contains a kernel of truth, maybe a whole stalk of veracity. But I find it quite interesting.
As a preface, some authors (creators) have accused AI companies of using their content to train AI models. One can pump the name “Esther Mahlangu,” into Google’s art making machine Namo Alabama or whatever and get outputs that look like Ms. Mahlangu’s art. How did Google’s smart software see Ms. Mahlangu’s geometric figures? I think Google must have ingested them, added metadata, decomposed them into Gemini building blocks, and happily and without much worry output facsimiles or what looked like Ms. Mahlangu’s designs.
A senior manager at a big tech company with a tight grip on a number of markets is having a hissy fit. The person in the playpen believes that a corrupt, criminal element is taking his toys. The senior executive’s assistant finds the scene as laughable as I do. Thanks, Venice.ai. Good enough. Not a cartoon. No toys in the air. But I don’t expect much from AI.
The operative work is “copy.” Maybe it should be called “harvesting” or “spidering”? From my point of view, Google has copied the intellectual spark and motif of her original and imaginative work. I am not sure Google is in the original business unless it comes to claims about its quantum supremacy. I grant that seems like a creative bit of wordsmithing. But I keep coming back to taking, using, and outputting of zeros and ones that are most just copies.
Now, what about this write up: “Google: Hackers Are Trying to ‘Clone’ Gemini for Cyberattacks”? The article states:
As the tech industry races to develop new AI models, Google alleges that “private sector” entities have been trying to reverse-engineer its Gemini chatbot by bombarding it with prompts intended to leak its secrets… But the company says it “observed and mitigated frequent model extraction attacks from private sector entities all over the world and researchers seeking to clone proprietary logic.” Model extraction isn’t your typical hacker-led “break-in.” Instead of exploiting a software glitch or infiltrating a corporate network, these attacks leverage legitimate access via Gemini’s API, which Google sells to software developers who want to build their apps around the chatbot.
The write up includes a diagram of how Google was “compromised”. (Didn’t Google buy Mandiant to deal with simple black box copying? Doesn’t Google use its own AI to defend its AI?) The write up states:
“This activity effectively represents a form of intellectual property (IP) theft,” Google alleges.
Google does not want its AI wizardry copied. I would remind people about these three examples related to Google’s unauthorized use of published information:
- Authors Guild v. Google, Inc. (Google Books)
- Perfect 10, Inc. v. Google, Inc.
- Oracle America, Inc. v. Google LLC
Google won each case because its taking was fair use. I am no lawyer. It looks to me that Google’s money, legal resources, and its ability to suggest that a nifty Google mouse pad might appear from a Googler’s briefcase helped it prevail. I just find it amusing that Google is miffed at people who are copying its stuff.
The legal eagles will take flight. Destination: Court rooms. Argument: You cannot do what we do.
Stephen E Arnold, February 18, 2026
Comments
Got something to say?
-
- Subscribe to Beyond Search
Feature archive
News archive
-
