What Meta’s Latest Moves Could Mean for the AI Persuasion Equation

Meta's two recent announcements should get every marketer’s attention.

On the surface, they look unrelated.

First announced on June 23 and officially launched globally on September 15-16 during IAB's inaugural Global Creator Week, the Meta Creator Marketing Hub combines the Creator Marketplace and Partnership Ads Hub into a single workflow, from finding creators to identifying brand-relevant content and turning it into Partnership Ads.

Muse is on the other side. Meta’s personal AI assistant is intended to act for users by using its own browser, linking to various services and carrying out tasks, such as making purchases with the user’s approval. Muse was launched Sept 8 in the US, with Canada added Sept 18.

If you look at each one on its own, they’re interesting product launches. If you consider them side by side, a greater whole then appears.

Meta is developing AI infrastructure on both sides of the persuasion equation.

On the one hand, AI enables brands to identify and expand the human voices that influence consumers. On the other hand, AI agents are now taking action on behalf of consumers.

That could create a very different marketing environment.

The Dual AI Persuasion Loop

For most of the digital era, we have assumed humans do the navigating.

We search, scroll through options, watch videos, click links, compare products, read reviews, go to websites, and in the end some of us make a purchase.

Consumer-facing AI agents may challenge that assumption.

I think the emerging journey could increasingly look something like this:

Brand Strategy → Creator Marketing → Creator Signals & Content → Evidence & Validation → AI Agent → Consumer Choice

This is not Meta’s stated architecture. It is my interpretation of where these developments could lead and how consumer tech brands may be thinking about it.

Creator content is just one indicator, alongside reviews, product specifications, pricing, availability, reputation, and third-party comments.

An AI agent could increasingly interpret some of those signals before presenting options or taking action.

Marketing may increasingly need to work beyond human attention.

Marketers will increasingly have to consider whether AI agents can find, understand, and trust evidence about their brand.

Industrialising Human Influence

There’s always been a rather interesting contradiction about creator marketing.

Its power comes from the fact that it is human.

What makes scaling difficult is also its human element.

Finding creators, assessing their suitability, handling outreach and permissions, getting their content ready, and linking organic success with paid media can be complicated.

The Meta Creator Marketing Hub makes the process more systematic.

Meta described the Hub as combining the Creator Marketplace with the Partnership Ads Hub to create a workflow that takes you from finding creators to identifying brand-relevant content and turning it into Partnership Ads. The Hub is also incorporating Facebook creators into the Creator Marketplace, together with more than five million Instagram creator profiles.

The important thing is that it is not just another tool for creators.

This is the industrial-scale application of human influence.

What was once:

Manual discovery → Outreach → Creator post → Campaign report

Increasingly becomes:

AI-assisted discovery → Collaboration → Permissioned content → Paid amplification → Measurement

The impact of humans becomes easier to scale.

At the same time, Meta is launching a technology that could alter the situation on the other side.

Muse Could Sit Between Persuasion and Choice

Imagine I need a carry-on suitcase for a three-day business trip.

Today, I might use a search engine, visit marketplaces, watch creator reviews, compare websites and eventually buy.

Tomorrow, I could simply say:

Get me a lightweight carry-on that costs under $300 and has good reviews with delivery before Friday.

How much of the journey between that request and the eventual purchase will I actually need to navigate myself?

Muse already has its own browser and can perform tasks across connected services. It can fill forms, negotiate on a user’s behalf and make purchases after getting approval.

Meta is also going deeper into commerce.

Muse launched with dozens of partners and access to the full Shopify catalogue. Meta is adding Walmart, Best Buy, Sephora, Wayfair and other retailers, as well as Shop Pay and PayPal for payments.

But not everyone is opening the door just yet. It’s an interesting development to watch because the emerging tension may not simply be about which AI agent is smartest.

It could be a matter of who controls access to customers, the product catalogue, and ultimately the transaction.

This means that marketers now have a new question regarding distribution.

If more and more consumers use agents to find products, then being accessible to those agents could become at least as important as appearing in search engines or on marketplaces.

One AI Extends Creator Influence. Another May Interpret It.

It is only when the two announcements are considered together that they become rather more interesting.

The brand uses the Creator Marketing Hub to find relevant creators.

Creators create content in order to build awareness, show their products, and generate third-party signals.

The brand promotes high-quality content by using Partnership Ads.

Then a consumer asks an AI agent to help them decide between products.

Another system now joins the journey. It could start by assessing product information, reviews, prices, availability, reputation and other indicators before displaying the various options.

It might also involve content created by the creator.

One AI could extend creator influence, while another AI might increasingly act as a mediator in decision-making.

That is a very different marketing environment from what we have been used to in the digital space.

Human Creators May Become More Valuable, Not Less

It is tempting to think that the greater the amount of AI the less human involvement there will be.

I believe the opposite may occur.

AI agents need signals they can interpret.

A company’s claim that its product is excellent is just a claim.

Customers saying it is excellent on their own is another indication.

A reliable, credible creator who shows off, tests, or explains the product provides another.

It doesn’t mean that AI agents will automatically consider creator content to be authoritative.

Creator marketing might become more than a means of generating attention.

It can add to the wider body of evidence concerning a brand.

The question therefore changes from:

"How many people saw this?"

to:

Can this content provide credible evidence for why one should choose us?

From Share of Attention to Share of Recommendation

For decades, marketers have competed for share of voice, share of search, impressions, and clicks.

Agentic interfaces introduce another interesting concept:

Share of Recommendation.

It matters that your product consistently comes up when someone asks an AI agent for five products.

If your brand doesn’t appear, understanding why becomes a marketing problem.

Is your product information clear?

Is evidence available to support your claims?

Is the pricing and availability correct?

Can an AI system tell what it is that actually makes you different?

That’s why structured product information, reviews, content from credible third parties, and useful demonstrations by creators could increasingly become part of the marketing infrastructure.

Brand Still Matters

This does not imply that brand building is any less important.

Consider the difference between:

"Buy me XYZ brand running shoes."

and:

"Find me good running shoes for travelling that I can also use in the gym."

The initial instruction reflects years of accumulated brand preference.

The second delegates much more of the decision.

Marketers now need to be successful in both situations.

Create a level of preference so that people start by asking for you directly by name.

And create sufficient evidence so that an agent will be confident about bringing you forward when they don’t.

Trust Becomes Part of the Architecture

Marketers shouldn’t overlook another dimension: who paid for the influence?

Meta already requires creators to disclose commercial relationships with brands through its Paid Partnership label. Partnership Ads also require the creator’s permission before a brand can amplify their content.

That distinction becomes even more important in an agentic world.

If an AI agent is evaluating product information, reviews and creator content, should it also understand whether a creator was paid, received a free product, earns an affiliate commission, or expressed an independent opinion?

And if brands eventually pay for visibility within AI-mediated recommendations, how should that commercial influence be disclosed?

The challenge may no longer be simply making advertising transparent to people.

It may also be about making the provenance of influence understandable to AI.

As creator marketing becomes easier to scale and AI agents play a greater role in discovery and choice, permissions, disclosure, attribution and commercial relationships become part of the information architecture.

Trust is no longer just a brand value. It could become a machine-readable signal.

What Should Marketers Do Now?

These are early days, but a chance for Marketers to get ahead of these conversations.

Muse is currently rolling out in the US, giving us an early glimpse of how this model could evolve as the ecosystem develops and potentially expands to other markets.

The success of consumer AI agents will depend on several things: whether consumers trust them to act on their behalf, whether ecosystem partners support them, and whether the experience proves genuinely useful.

Having seen consumer tech evolve over three decades, I think this will become one way consumers use AI. If agents can genuinely remove friction while earning consumer trust, the model has a strong reason to gain traction.

These are the considerations I would evaluate as a Marketer.

• Can AI systems accurately understand our products and services?

• Is our product information up to date, well-organised, and consistent?

• What independent evidence supports our product or service claims? Reviews, testimonials, UGC, etc.

• Are creators demonstrating utility, rather than simply endorsing us?

• Are creators' permissions and usage rights properly managed?

• Can the new agent ecosystem really get access to our products and information?

• What would occur if discovery, comparison and transaction were to take place outside of our website?

• Can we start understanding our share of recommendation, not just our share of attention?

Meta’s two announcements do provide an interesting glimpse of how marketing and commerce may evolve.  

For the last two decades, digital marketing has largely asked:

How do we get found, noticed and clicked?

The next era may add another question:

How do we build a brand that humans prefer, creators credibly advocate for, and AI agents have enough evidence to recommend?

And perhaps the simplest test is this:

If an AI agent evaluated your category today, would it understand why your brand deserves to be in the consideration set?