You’d be forgiven for remembering that Meta also has an AI, Meta AI. Or for not knowing at all. There’s a crowded field of front-runners in the AI space: OpenAI’s ChatGPT, Anthropic’s Claude, Google’s Gemini, Perplexity, Microsoft’s Copilot, and xAI’s Grok.
In a matter of days, Meta will begin feeding users’ interactions with Meta AI into the algorithms that control what content Meta shows to you, including reels, posts, and ads.
Meta began spreading the word about changes to Meta AI back on October 1, 2025, but the changes haven’t yet gone live. The new policy takes effect on Tuesday, December 16, 2025.
what changes
“We will start personalizing content and ad recommendations on our platforms based on people’s interactions with our generative AI features,” Meta posted in its October 1, 2025, press release. Meta AI will use both text interactions and voice chats as fodder for further content personalization.
“Your interactions with content on Facebook and Instagram have long shaped what appears in your feed,” Meta elaborated. “Just like other personalized services, we tailor the ads and content you see based on your activity, ensuring that your experience evolves as your interests change… Soon, interactions with AIs will be another signal we use to improve people’s experience.”
They’re not wrong, but there’s something extra unsettling when it comes to retaining, even temporarily, and using voice commands. Aside from in-app notifications, Meta has begun sending out emails to holders of Meta accounts to notify them of the imminent changes to its privacy policy.
“With tools like Ads Preferences and other feed controls, you can continue to adjust the content and ads you’re seeing at any time,” reads the email notification Meta sent out to users. Whether that means users will be able to turn off Meta AI’s scooping up of interactions or not, we don’t know due to the vague language. I suppose we’ll all find out in a few days.
The post Meta Is Going to Start Using Your Interactions to Train Its AI appeared first on VICE.
