Google Will Record Everything Users Post Online to Train Its AI Products

It's open season on public posts as Google announces its new privacy policy

Demis Hassabis, co-founder of artificial intelligence startup DeepMind, in Seoul, South Korea, on March 15, 2016. Hassabis now heads Google's AI unit, which recently merged its own AI lab, Google Brain, with DeepMind.Jeon Heon-Kyun-Pool/Getty Images

Google will record everything people post online in order to train its artificial intelligence products.

On July 1, Google amended its privacy policy to allow it to scrape comments that posters put on the internet, to help it to hone its AI tools.

The tech company’s plan to harvest and harness online public data is raising new privacy concerns.

Google’s previous user policy stated that publicly available information would only be scraped to help train its “language models” for Google Translate.

The U.S. tech giant’s history of changes to its user policy statement is open to the public.

There are also concerns that advanced AI technology will be used to steal intellectual property and eliminate several professions done by humans, along with violating user privacy.

Google Scraping User Data To Improve Its Bard AI System

After a delayed and troubled release, Google’s AI chatbot, Bard, which launched months after Open AI’s ChatGPT release, quickly caught up with its key rival.

Google and OpenAI have already scraped vast portions of the internet to fuel their AI bots’ algorithms.

So far, the tech giant seems to have shifted its data collection focus from language to AI models, and mentioned Bard and Cloud AI for the first time in its updated service terms.

Google will keep and read any public comments from now on, with some of them being retained for chatbot training.

“Google uses information to improve our services and to develop new products, features and technologies that benefit our users and the public,” reads the new Google policy.

“For example, we use publicly available information to help train Google’s AI models and build products and features like Google Translate, Bard, and Cloud AI capabilities.”

The new policy update did not make any changes that would directly impact user experience or Google products.

Nonetheless, the amendments to Google’s previously stated user policy show that it is investing heavily in its AI program and that skimming general search behavior could be a significant factor in its continued development.

The California-based tech company has also announced a new AI-based search system, known as the Search Generative Experience (SGE), as part of its new lineup of AI products.

Google is preemptively informing users of its future AI plans, by dropping hints that it has been developing new AI products for its systems including shopping, Google Lens features, and a text-to-music generator.

AI Bots Already Sparking Legal Challenges And Social Media Company Resistance

The data scraping policies led to a class action lawsuit involving the matter last week, reported Gizmodo.

The plaintiffs accused Open AI of stealing “essentially every piece of data exchanged on the internet it could take” without credit, consent, or compensation.

Google’s policy changes remain in legally murky waters, as the courts begin to wrestle with new copyright issues over AI tech—issues that were unheard of not long ago.

Most privacy policies restrict companies to collecting data that users provide directly, but Google’s new policy allows it to use any information people post publicly online.

Alphabet, Google’s parent company, warned its own employees about the security risks of using chatbots in June, while releasing its own Secure AI Framework to protect users from AI cyber threats.

Meanwhile, Twitter and Reddit reacted negatively to the AI data scraping controversy, making major changes to restrict access to their platforms from bots.

Both companies turned off free access to their API’s—application programming interfaces—which allowed users and AI bots to download large quantities of data from posts.

Last week, Twitter limited the number of tweets users were allowed to look at per day, causing chaos for the social media giant as users encountered error messages and limited access to tweets.

CEO Elon Musk said the changes were a necessary response to “extreme levels of data scraping and system manipulation.”

Twitter also began requiring an account to view tweets, then quietly removed the requirement earlier this week.

Google did not immediately respond to The Epoch Times’ request for comment.