According to a recent report, Meta has been quietly working on a web-crawling search engine for its chatbot, Meta AI, aiming to provide up-to-date, conversational responses on current events. The company has also teamed up with Reuters to ensure its AI can tackle news-related questions.

Meta has been indexing the web for around eight months, with the project reportedly led by senior engineering manager Xueyuan Su. Although the company hinted at its web-crawling tech earlier this year, framing it as part of "training AI models or improving products," it stopped short of confirming a full search backend.

This move is said to be Meta’s response to feeling "stung" by its reliance on other tech giants, particularly after Apple's App Tracking Transparency (ATT) feature disrupted Meta's ad revenue.

Since ATT launched in 2021, Meta estimates it has lost over $10 billion in ad earnings. The company was so rattled by the feature that it faced a lawsuit for allegedly finding ways around ATT's rules.

CEO Mark Zuckerberg is reportedly pushing for Meta to stand on its own, aiming to avoid a similar dependency mess if Google or Microsoft were to restrict its access to web searches. It's not clear if Meta currently pays either company for access, but the goal seems to be full independence.

Meta's big AI investments seem to be gaining serious traction. In August, Zuckerberg shared that Meta AI now has over 185 million active users every week and more than 400 million every month.