Desperate Google is buying trash-code from Android developers to save its AI
When your multi-billion-dollar AI models can't code their way out of a wet paper bag, what do you do? If you are Google, you quietly slide into developers' DMs with bags of cash, begging for their dusty old side projects.
Journalists at 404 Media caught Google red-handed sending out secret emails to successful Google Play developers. The corporate giant is offering to buy active production codebases, old prototypes, and forgotten side projects that are just gathering dust in digital archives.
Because nothing says "we are leading the AI revolution" quite like dumpster-diving through a random indie developer's college homework. The targeted developers are being approached individually, keeping the whole scheme safely away from public eyes.
In the leaked emails, Google calls this sketchy initiative a "confidential content offer pilot." They promise developers a nice chunk of "additional revenue" while letting them keep 100% of their intellectual property. The pitch claims this real-world, production-tested code will simply "improve developer tools," but the links in the email point directly to Google's AI partnership programs.
The reality behind this sudden corporate generosity is simple. Google is visibly trailing behind rivals like Anthropic with its shiny Claude Code and OpenAI's programming models. It turns out scraping open-source GitHub repositories didn't give Gemini enough brainpower to write flawless code, forcing the tech giant to buy private data. The exact number of targeted developers and the size of the payouts remain strictly confidential.
The tech industry has officially entered the era of data desperation. When scraping the public internet is no longer enough, trillion-dollar corporations will happily pay to rummage through old digital trash cans just to gain a slight edge in the AI arms race.
Source: 404 Media
Comments
This is where the magic happens: AI reads your discussion and rewrites the article based on the most interesting comments. Each strong comment adds points to the meter below. Once the meter is full, the article updates live — no page reload needed.