Meta’s latest AI launch adds another layer to how people may discover brands, products and services across its platforms.

Muse, launched on 8 September, is Meta’s new personal AI agent. It can carry out tasks across apps, help with planning and progress longer-term goals. People can use Muse through its own app or directly inside WhatsApp.

Muse sits alongside a much broader expansion of AI across Facebook, Instagram, WhatsApp, Messenger and Threads. Meta AI can answer questions, surface recommendations and increasingly draw on content from across Meta’s platforms. Business Agents are meanwhile giving brands their own AI layer for handling customer conversations.

Those pieces are beginning to connect.

Meta has spent years using AI to decide which content people see. AI is now starting to influence which brands people discover, which products they consider and what they do next.

Muse Makes AI Discovery More Actionable

Meta describes Muse as an agent that can send emails, book travel, organise plans and operate across connected services on the user’s behalf. It can also draw on activity elsewhere in Meta’s ecosystem. One example involves Muse taking a recipe Reel saved on Instagram and turning it into a shopping list.

That starts to extend Meta AI beyond recommendation and into action.

Someone planning a trip might move from asking for destination ideas, to identifying accommodation, to organising parts of the trip. Someone looking for a product could move from inspiration towards comparison and purchase without treating each stage as a separate digital journey.

For brands, visibility inside AI-generated recommendations becomes more commercially significant when the same system can help progress the decision.

Discovery Is Moving Beyond The Feed

Facebook and Instagram have already moved a long way from feeds built mainly around the accounts people choose to follow.

Reels accelerated the shift towards content selected according to interests, predicted behaviour and engagement patterns. Meta AI adds a more active form of discovery on top of that.

Instead of waiting for something relevant to appear, users can ask for it.

Meta’s Muse Spark model, which powers Meta AI and Muse, can surface recommendations from Reels and other sources. Meta has also said the model will increasingly cite recommendations and content shared across Instagram, Facebook and Threads.

Shopping offers one indication of where that may lead. Meta AI can search Facebook Marketplace alongside products from elsewhere online, while users can reference particular brands or creators and explore their public content.

A brand can therefore be surfaced in more than one way. It can appear because an algorithm selects its content for the feed, or because an AI system decides the brand, product or piece of content is relevant to a question.

Brand Content Gets A Second Job

As Meta AI draws more heavily on public content, the role of brand content starts to widen.

Posts, Reels, creator content and business information are no longer valuable only for the audiences they reach directly. They can also help Meta’s systems build a clearer understanding of what a brand does, what it offers and when it may be relevant to a user.

Meta has already indicated that Muse Spark can use recommendations and content shared across Instagram, Facebook and Threads. Its AI shopping tools can also explore the public content of specific brands and creators.

That creates a second layer of value around content marketing. A Reel may still be judged on reach, engagement or conversion, but it could also contribute to how accurately Meta AI interprets and surfaces the brand later.

Similar patterns are emerging around Google AI Overviews, ChatGPT and other AI discovery environments. Meta brings social content, creator activity and community discussion much more directly into that picture.

Meta Has A Different Discovery Advantage

Meta’s position in AI discovery differs from that of a standalone assistant.

ChatGPT has deep context from what users ask it. Google has enormous visibility into search intent and behaviour across its services.

Meta sits across years of signals around what people watch, follow, click, share and engage with on Facebook and Instagram, alongside messaging and commercial activity elsewhere in its ecosystem.

AI interactions are becoming another signal. Meta began using people’s interactions with its AI products to help personalise content and advertising recommendations in late 2025.

Muse adds further context because it is designed to learn from conversations and remember what matters to an individual over time.

Meta separates some Muse data from advertising systems, so it would be wrong to assume every behavioural signal feeds directly into every AI recommendation. Even so, the company already has a substantial personalisation infrastructure around which to build AI discovery.

A recommendation based only on a question is one thing. A recommendation shaped by a wider understanding of interests, behaviour and current intent is potentially much more useful.

Business Agents Connect Discovery With Engagement

Meta’s Business Agents bring the brand side of the journey into the same picture.

Businesses can deploy AI across WhatsApp, Messenger and Instagram to answer questions, recommend products, book appointments, qualify leads and, in some cases, help complete sales.

More than one million businesses were already using a Meta Business Agent across WhatsApp and Messenger when Meta announced the wider rollout.

Meta is also making those businesses easier to find. WhatsApp users will increasingly be able to discover businesses with a Meta Business Agent through search, contact cards and recommendations shared between users.

Muse works for the consumer. Business Agents work for the brand.

A user could discover a relevant business through Meta AI or Muse, move into a conversation with that company’s Business Agent, ask questions, receive recommendations and progress towards a booking, lead or purchase.

The boundaries between discovery, engagement and conversion start to become much less distinct.

Meta AI Visibility Is Still Developing

The measurement gap around Meta AI is starting to narrow, but it is still well behind the visibility marketers have around search, social feeds and paid media.

Profound already offers Meta AI monitoring through the meta.ai web experience, giving brands a way to track how they appear in AI-generated responses. Other AI visibility platforms will almost certainly expand in the same direction as usage grows and client demand follows.

Coverage across Meta’s wider ecosystem is still much less mature.

A brand may have limited visibility into how often it is recommended inside Facebook, Instagram or WhatsApp, what information influenced the answer, which competitors appeared instead or how results vary between users.

That is likely to change.

As Meta AI adoption grows, third-party tools will have a strong incentive to extend their coverage. Meta itself is also likely to improve APIs, reporting and access for partners and measurement providers, much as other major platforms have done as new discovery behaviours became commercially significant.

For now, the market is in an awkward middle phase. Some monitoring exists, but Meta’s AI experience is evolving faster than the measurement layer around it.

Social Strategy Will Need To Widen

Meta AI adds another job for social content.

Posts, Reels, creator activity, business information and community discussion are still being created for reach, engagement and conversion. Increasingly, they may also help shape how Meta’s AI systems understand and recommend a brand.

That broadens the role of social strategy.

Teams will need to think not only about what performs in feeds, but also about whether their overall presence gives Meta enough useful, consistent information to understand what the brand offers, where it is relevant and how it differs from competitors.

Creator partnerships may matter here too. Public creator content can add another layer of context around products, experiences and brand perception, particularly as Meta AI draws more heavily on content from across Instagram, Facebook and Threads.

Social discovery has already moved well beyond followers. AI now adds a conversational and increasingly agentic layer on top.

Muse can help people act on what they discover. Business Agents can continue the interaction once a brand enters consideration.

For social teams, content strategy is gradually becoming part of AI discovery strategy as well as feed strategy.