Google Updates Privacy Policy to “Scrape Everything You Post Online” for AI Development

Google recently updated its privacy policy, granting itself the authority to scrape virtually all online posts to improve its artificial intelligence (AI) capabilities. The broad language suggests that if Google can access your words online, it’s safe to assume that they now belong to the company. As users of the internet, we should be prepared that our posts may be used as training data for a chatbot or another AI tool in the Google ecosystem.

According to the revised policy, “Google uses information to improve our services and to develop new products, features, and technologies that benefit our users and the public.” Google has clarified that it will use publicly available data to train its AI models and to develop products and features, such as Google Translate, Bard, and Cloud AI capabilities.

Changes to Google’s terms of service can be tracked via its maintained history, which now includes the updated policy language. While the company previously stated that data would be used “for language models,” the new wording emphasizes “AI models.” In addition, Google Translate, Bard, and Cloud AI are explicitly mentioned as beneficiaries of the data harvested.

This clause is highly unusual for a privacy policy. Traditionally, these policies detail how a company uses information that users post on their own services. However, Google’s updated policy implies that the company reserves the right to harness data posted on any part of the public web, essentially treating the entire internet as its AI playground.

The new policy raises several interesting and complex privacy issues. Although it is generally understood that public posts are public, the lines become blurred when considering the extent to which this information can be used. Bard and ChatGPT, for instance, could have processed your long-forgotten blog posts or 15-year-old restaurant reviews, and may be repurposing your words in unpredictable and hard-to-understand ways.

Another complex issue in the post-ChatGPT world is the question of where these data-hungry chatbots source their information. Both Google and OpenAI have been known to scrape vast portions of the internet to fuel their AI models. The legality of this practice remains murky, and the coming years will likely see courts grappling with copyright issues that were unimaginable until recently.

The repercussions of this AI data sourcing are already being felt by consumers and platforms alike. In a bid to prevent their intellectual property from being harvested, social media giants Twitter and Reddit have made controversial changes to lockdown their platforms, cutting off free access to their APIs. These changes have disrupted third-party tools that many people used to access these sites, causing widespread discontent.

Elon Musk, the influential tech magnate, has been particularly vocal about web scraping. He recently blamed a number of Twitter’s technical issues on its need to prevent data scraping and system manipulation, despite most IT experts attributing these problems to internal mismanagement or incompetence.

Reddit has been hard hit by these API changes, sparking a mass protest from unpaid moderators who relied on third-party tools for managing the forums. The protests resulted in a large part of Reddit effectively being shut down, a move that will likely have permanent repercussions.

In conclusion, Google’s new privacy policy is a significant departure from the norm, turning the whole internet into a playground for its AI models. This has triggered interesting debates on privacy, legality, and the control platforms have over their own data. As the dust continues to settle, we can expect significant changes to the online landscape, with users, tech companies, and legal systems needing to adapt to this new reality.