Reuters reported on October 1:
Meta Platforms (META. O), opens new chapter, reporting Wednesday, November 12, 2025 that it will begin using human interactions with its generative AI tools to personalize content and ads on apps like Facebook and Instagram starting December 16.
Users will be notified of the changes from October 7 and will not have an option to opt out, the social media giant said, though the update only applies to those who use Meta AI.
Description of the proces
Meta will use AI conversations (text and voice) as additional input for ad and content personalization. (Reuters)In short:
- Users talk to a Meta AI assistant (in apps like Meta AI). (The Keyword)
- Meta analyzes the topics from those conversations (insofar as they do not fall under “sensitive topics” such as religion, health, etc.). (Reuters)
- These conversational insights are combined with other signals (such as likes, following behavior, purchase history) to determine which ads or content a user receives. (Reuters)
- Based on this, ads can be shown that are more in line with what the user is apparently looking for or discussing — for example, “If someone talks about walking, you’ll see ads for hiking shoes later.” (The Wall Street Journal)
What can an advertiser do with it?
As an advertiser, you can benefit from this development in several ways. Here are concrete actions and points of attention:
- Create more relevant adsMake sure your ads cater to what users think/are looking for , not just what you used to think they wanted. Because Meta gets more context, you can become more relevant.
- Use conversational insights as inspirationThink about what questions or problems your target audience has: “What are they talking about?”, “What are they trying to solve?”, “What terms are they using?”
Develop ad creatives, headlines, and call-to-actions based on them. - Optimize landing pages and contentMake sure that after the click, the experience makes sense: if your ad is the result of something someone discussed (e.g. hiking), then the landing page should connect (hiking shoes, trails, tips). This increases conversion.
- Segmentation and audience alignment
Suppose Meta’s system recognizes users who talk about a particular topic (e.g., “dog walking,” “sustainable travel”). You can set up campaigns that specifically cater to those interest clusters. - Measure and test proactivelyBecause the technology is new, it is smart to run A/B tests: which ad works better if it responds to “question/conversation topic” vs standard ad? Track KPIs like CTR, conversion, cost per acquisition.
- Be mindful of privacy and regulationImportant: Meta says conversations about sensitive topics are not used for targeting. (Tom’s Guide)So you have to not focus on themes such as health, religion, ethnicity if you expect Meta to use that data. Also: in some regions (EU, UK, South Korea) stricter rules apply or the change applies later. (Reuters)
Practical steps for your business
- Map out what your target audience often talks about: FAQs, chats, social media comments.
- Test ad variations that address these questions/conversations (question-driven headlines, “Do you have this problem too?” promotion).
- Set up a conversion measurement so you can see if this new approach is actually working better.
- Keep ad creation lean and keep iterating fast: the target audience changes, the AI context changes.
- Work with your agency or internal team to make sure data and content are aligned with that (so that when someone clicks through, the experience connects).
