Right now, what’s as important as the AI powering our world is how AI models are trained. We all know that these AI models need to be trained on a ton of data. The issue is that much of this data was taken without permission. According to a new report, the EU has launched a probe into Google to see if it illegally obtained data to train its model.
The subject of data privacy is pretty hot, as major corporations continue to ignore basic privacy laws. There’s no telling how much data these companies obtained illegally. For example, Meta just admitted to scraping data from Australians as far back as 2007 (back when it was still called Facebook).
The EU launched a probe into Google over its AI model
We’re all pretty sure that most of the data powering Gemini was illegally obtained, but this probe isn’t about Gemini, as Google has other models. The Ireland-based DPC (Data Protection Commission) expressed concern over how Google trained its Pathways Language Model 2 or PaALM 2. This is a foundation model introduced back in May 2023.
The commission wants to know if Google respected the privacy of EU citizens when training the model. “This statutory inquiry forms part of the wider efforts of the DPC, working in conjunction with its EU/EEA (European Economic Area) peer regulators, in regulating the processing of the personal data of EU/EEA data subjects in the development of AI models and systems,” it stated.
This is a valid concern. As time goes on, these models get stronger as the companies behind them race to make them the best on the market. However, the issue is that companies are scooping up more data as time goes on. We had data protection laws in effect, but it wasn’t until the generative AI boom that we realized just how much data was being scraped on the daily.
At the moment, this is just an inquiry, so it’s not an indication of a lawsuit or any legal action. However, if it happens that Google illegally obtained any information, then we can expect legal action to follow.
2024-09-12 15:06:49