Meta has announced a major update to how it will personalise content and advertising across Facebook, Instagram, Threads and WhatsApp. Since October, the company has begun notifying users that it will start using people’s interactions with Meta AI to refine its recommendation systems and improve ad targeting from 16 December 2025.

The change means that any text or voice conversations with Meta AI could influence the content and ads users see across Meta’s family of apps. For marketers, this marks one of the most significant shifts in Meta’s data ecosystem since the introduction of interest-based targeting more than a decade ago.

For users, it is another reminder that generative AI is not just a utility but a data layer that feeds the algorithms behind the world’s biggest advertising engine.

How It Works

Meta says more than one billion people now use Meta AI each month. These interactions, whether asking for travel tips, generating images or chatting about hobbies, will soon be used as new signals for personalisation.

In practice, this means that conversations with Meta AI will be treated similarly to how the company currently uses likes, follows and engagement data to refine what appears in users’ feeds.

If, for example, a user chats with Meta AI about hiking, the system will infer an interest in hiking. As Meta explains:

“If you chat with Meta AI about hiking, we may learn that you’re interested in hiking, just as we would if you posted a reel about hiking or liked a hiking-related page. As a result, you might start seeing recommendations for hiking groups, posts from friends about trails, or ads for hiking boots.”

This approach effectively folds conversational AI into Meta’s core ad-targeting infrastructure. The company stresses that sensitive topics such as religion, sexual orientation, political beliefs and health will be excluded from this process.

A New Layer of Data Intelligence

For marketers, the implications are clear. Meta is expanding its understanding of user intent beyond traditional engagement signals to include AI-driven insights derived from private interactions.

This could deliver a much richer, real-time view of consumer interests. Where someone’s browsing or posting behaviour might offer only indirect signals, AI conversations provide a window into what people are actively thinking about, planning or seeking.

A user who asks Meta AI to recommend budget hotels in Bali is sending a far stronger intent signal than someone who simply likes a travel photo. In theory, that could make ad placements more relevant and effective, especially for categories driven by timing or context such as travel, retail or entertainment.

It also highlights Meta’s strategic goal of making its AI assistant an integral part of the commercial funnel. If users begin using Meta AI to search for products or seek recommendations, Meta gains access to a new, high-quality dataset that goes far beyond passive behaviour.

Rebuilding Relevance After Privacy Restrictions

Since Apple’s App Tracking Transparency policy began limiting third-party data access in 2021, Meta has been forced to rethink how it gathers and interprets signals. The shift towards first-party conversational data is the next logical step.

Rather than relying on external tracking, Meta is doubling down on data generated within its own platforms. Conversations with Meta AI are firmly within Meta’s ecosystem, meaning they can be used to rebuild the precision of its ad targeting without breaching new privacy constraints.

This also helps Meta future-proof its ad products against further privacy regulation, which is tightening across the US, EU and APAC markets.

User Control and Transparency

Meta began notifying users about the update via in-app alerts and email in early October, several weeks before the policy takes effect. Users can manage their preferences through Ads Preferences and feed controls, allowing them to see why specific ads are shown or to adjust relevance settings.

Interactions on accounts not linked within Meta’s Accounts Center, such as using Meta AI in WhatsApp without linking it to a Facebook or Instagram account, will not be used for cross-platform targeting.

Meta also reiterates that microphone use will only occur with explicit permission when using voice features, and that an indicator light will signal when audio input is active.

While these assurances provide some transparency, the update still represents a significant step forward in Meta’s use of conversational data. For brands, this means a new layer of personalisation potential. For users, it raises familiar questions about how much of their digital behaviour feeds into the ad ecosystem.

What It Means for Marketers

For marketers and CMOs, this change has both promise and complexity.

1. Richer intent data. AI conversations offer direct, context-rich insight into user preferences. Unlike likes or follows, chats reveal specific needs and motivations. Campaigns could be built around more immediate consumer intent, improving conversion potential and reducing wasted impressions.

2. New targeting opportunities. If Meta opens up anonymised or aggregated AI-based segments within Ads Manager, advertisers could target users based on conversational themes such as travel planning, product research or wellness goals. This would enable highly granular creative and messaging strategies.

3. Smarter creative testing. As conversational signals evolve quickly, creative performance testing may need to become more dynamic. Marketers will need to adapt campaigns in near real time to align with changing intent patterns drawn from AI activity.

4. Responsible data use. Marketers must also tread carefully. While Meta has promised to exclude sensitive data, users may not fully understand how their AI interactions are used. Transparency in messaging and respect for consent will be crucial to maintaining trust.

The Broader Shift Toward AI-Driven Targeting

Meta’s move underscores a wider trend across the advertising industry. AI conversations are becoming a new frontier of behavioural data.

If chatbots and generative AI interfaces become the way people search, plan and shop, those interactions will increasingly shape how ads are delivered. Meta’s update foreshadows what is likely to happen across other AI ecosystems, from Google Gemini to Amazon’s Rufus and OpenAI’s ChatGPT.

Each platform is looking to integrate conversational intent into its advertising infrastructure, creating a feedback loop between AI assistance and commercial outcomes.

For CMOs, this means preparing for an era where conversation equals data. Understanding how to interpret and activate these signals while staying compliant with emerging data laws will be central to marketing effectiveness in the years ahead.

A Step Toward AI-Native Advertising

Meta’s integration of AI chat data is not just a feature update. It is a signpost toward the next phase of digital marketing. Ads will increasingly be informed not just by what people click or post, but by what they ask, say and imagine with AI.

If Meta succeeds in turning these conversational insights into tangible advertising performance gains, it could redefine how brands reach and resonate with audiences.

The line between conversation and commerce is blurring fast. For marketers, the opportunity and the responsibility is to ensure their use of AI-enhanced data delivers relevance, creativity and trust in equal measure.