Google could use public data for AI training, according to new policy

    Google can now use public data to help train and create AI products, according to new privacy policy changes(opens in a new tab).

    As of July 1, the tech giant’s newly adjusted policy(opens in a new tab) reads: “Google uses information to improve our services and to develop new products, features and technologies that benefit our users and the public. 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.”


    Reddit’s API protest just got even more NSFW

    Previously, the policy only stated that publicly available information could be used to help train Google “language models” and gave a single mention of Google Translate.

    While the update doesn’t change the user experience or directly impact Google products as of right now, the language adjustment signals that the company is leaning more heavily into its AI bid, and that the general public’s search behavior could be a significant factor in its continued development.

    Google has hinted at several entries into the AI space, including AI shopping experiences, Google Lens features, and even a text-to-music generator.

    While Google’s AI chatbot, Bard, launched to a less than enthusiastic response at first, it’s quickly caught up with other chatbots on the market. Google also announced an upcoming AI-based search known as the Search Generative Experience (SGE) to round out its lineup of AI offerings. Ironically, Google’s parent company, Alphabet, warned its own employees about the security risks of using chatbots just last month, and Google has released its own Secure AI Framework in an effort to enhance cybersecurity around AI threats.

    Broadly, concerns about privacy, intellectual property, and the impact of these models on human labor and creativity have plagued the introduction of new AI products. A class action lawsuit was brought against OpenAI(opens in a new tab), the maker of popular AI bot ChatGPT, just last month, claiming that the company stole “essentially every piece of data exchanged on the internet it could take” without notice, consent, or compensation.

    Some online harkened the Google update to the controversial ClearView AI(opens in a new tab), which built a law-enforcement grade facial recognition app by reportedly sourcing billions of facial images from social media sites and other platforms. In 2022, ClearView AI settled a lawsuit with the ACLU, banning the company’s ability to sell or give away access to its facial recognition database to private companies and individuals.

    For now, Google seems to be preemptively warning users of its future AI plans. Just be aware that your searches might be making an AI bot smarter.

    Read the full article here

    Recent Articles

    Related Stories

    Leave A Reply

    Please enter your comment!
    Please enter your name here

    Stay on op - Ge the daily news in your inbox