Google’s $10M Spirit Airlines Data Grab: The AI Training Grift Nobody’s Calling Out

Let’s talk about the hottest commodity this sweaty August: dead data. While everyone’s out on the Rockaways pretending not to think about work, Google just scooped up Spirit Airlines’ digital remains for a paltry $10 million—a price that wouldn’t cover a week of their ad fraud losses. Forget bankruptcy vultures and distressed asset funds; the real ghouls are the tech giants snapping up billions of decommissioned emails, chat logs, and operational code like it’s an end-of-summer garage sale.

According to Bloomberg Law, this so-called ‘trove’ includes 100 million emails and 500 million Microsoft Teams chats. You think Google cares about customer service logs from a defunct budget carrier? Get real. This is about AI training data. The gold rush isn’t for models that understand the present, it’s for models that can hallucinate the past—at scale, without privacy headaches. De-identified, yes, but let’s not pretend Google’s not salivating over the latent behavioral patterns and operational metadata buried in those files.

Meanwhile, ChatGPT’s parade of “miraculous” advertisers has done its own disappearing act. One minute, OpenAI is crowing about CPG brands lining up to dump cash into AI-powered ad units; the next, those same brands are gone faster than a Spirit flight refund. The lesson? Nobody wants to be the first to admit the ROI on these AI ad experiments is still a pixelated mirage.

Here’s the uncomfortable truth: the only people getting rich off dead data and vaporware ad platforms right now are the ones who know how to package, re-label, and resell digital trash to the next round of optimists. Don’t kid yourself—if you think your own company’s “de-identified” Slack logs aren’t the next loot in line, you haven’t been paying attention all summer.