The Publicis’ $3bn Liveramp deal signals shift in AI ‘Intelligence Layer’ ownership

Last week, news broke of Publicis’ $3bn LiveRamp deal, sparking industry questions over its implications for the wider marketing and data industry. 

This article explains how the acquisition signals a much bigger shift and that in the AI era, customer identity and intelligence infrastructure are becoming strategic enterprise control points.

The ‘customer graph’ is becoming the operating system for enterprise AI. That is the real story behind Publicis’ acquisition of LiveRamp

It’s not about consolidation, scale, or martech synergies. The question this deal forces every brand to answer is simpler and more consequential: who controls the intelligence layer your AI runs on?

For years, identity resolution and data collaboration were treated as plumbing — necessary, unglamorous, largely invisible. Publicis just paid $2.2 billion to make clear that they are now the foundation on which AI understands customers, makes decisions, and drives action across the enterprise. That is not infrastructure anymore. That is leverage.

Here is what most AI strategies get wrong: The model is not the moat

Foundation models are commoditising fast, and within a few years most enterprises will have access to roughly equivalent AI capabilities. The differentiation will not come from which model you run. It will come from what that model knows about your customers, and how accurately, durably, and quickly it can act on that intelligence.

In that environment, identity, first-party customer data, and interoperability become foundational enterprise infrastructure. The companies that win will be the ones capable of creating trusted customer context that AI systems can act on in real time.

That is the bet Publicis just made. And it should prompt every brand to ask whether they are making it too — or ceding that ground by default.

AI does not fix broken customer intelligence

There is still a tendency in the market to talk about AI as though the model itself is the differentiator. In practice, the differentiator is whether AI has access to accurate, durable, and real-time customer context.

If the underlying customer data is fragmented, duplicated, stale, or disconnected from durable identity, AI simply accelerates poor decisions. A recommendation engine becomes confidently wrong. Suppression fails. Media spend gets wasted targeting customers who have already converted. Personalisation becomes irrelevant because different systems hold conflicting versions of the same customer.

We already see this problem today. A customer purchases in-store but continues receiving acquisition ads for days because identity systems across retail, media, and CRM environments are disconnected. AI does not solve that problem. It amplifies it if the underlying customer graph is inaccurate.

That is why this acquisition matters beyond advertising technology.

The industry is shifting from systems of record to systems of decision and action. The companies pulling ahead are not necessarily the ones with the most data. They are the ones capable of continuously transforming customer signals into intelligent decisions in real time.

Durable identity is the moat. Not the model.

Brands should own the intelligence layer

LiveRamp built its business on neutrality. Sitting between advertisers, publishers, retailers, and agencies, it acted as trusted connective tissue — a Switzerland of identity. Publicis acquiring it changes that, regardless of what the integration roadmap says. You cannot be neutral infrastructure when you are owned by one of the largest buyers of media on the planet.

The conflict of interest is not subtle. Publicis now has a financial stake in how your identity data is resolved, governed, and operationalised — and a separate financial stake in winning your media budget. Those two incentives do not point in the same direction. Enterprises should not need to trust that they will be managed independently. They should not have to trust at all.

This is not a criticism of Publicis. It is a structural reality. And the right response for enterprises is not to hope for the best — it is to own the layer that makes the question irrelevant.

This will accelerate demand for independent customer intelligence layers that sit above activation and media ecosystems rather than inside them — particularly for enterprises operating across multiple agencies, retail media networks, cloud providers, and AI platforms simultaneously.

Most large enterprises do not want their customer intelligence strategy dependent on a single identity, media, or activation ecosystem. They want the flexibility to choose the partners and activation rails that best fit their business while maintaining ownership of the customer graph itself.

That becomes especially important as more enterprises build internal AI capabilities and deploy AI agents across marketing, service, and commerce. The intelligence layer powering those systems is becoming too strategic to fully outsource or embed inside a single activation ecosystem.

The black box problem gets harder in the AI era

There is a harder conversation underneath this deal that the industry has been reluctant to have. Black-box identity systems are increasingly incompatible with AI-driven enterprises.

As AI becomes more deeply embedded across marketing, commerce, and service operations, enterprises will demand greater visibility into how customer identity is resolved, governed, and operationalised.

Black-box identity systems become harder to justify when customer intelligence is powering automated decisioning at scale.

Brands increasingly want to understand how identity is being resolved, how customer data is governed, whether customer intelligence remains portable across systems, and how AI-driven decisions are being made from that data.

That pressure toward transparency and governance will only intensify as AI agents become more operationally embedded across the enterprise.

Interoperability will become a strategic advantage

One of the most important implications of this acquisition is that ecosystem flexibility itself is becoming strategically valuable.

The future will belong to enterprises that can create a durable customer intelligence foundation once and operationalise it everywhere.

That requires more than a customer database. It requires an enterprise-owned intelligence layer capable of interoperating across cloud environments, media ecosystems, retail networks, identity frameworks, and AI systems without forcing businesses into rigid architectures or closed ecosystems.

This acquisition will accelerate enterprise demand for warehouse-native and composable architectures. Brands that locked themselves into closed ecosystems are already feeling the cost. The ones building on open, portable foundations are better positioned for whatever comes next — whether that is a new AI platform, a new identity framework, or the next acquisition that reshapes the landscape.

The more fragmented the ecosystem becomes, the more valuable neutral orchestration layers become.

The next competitive battle will be won at the intelligence layer

The Publicis-LiveRamp deal is a signal, not an anomaly. The intelligence layer is being consolidated. The only question is whether your enterprise is building one you own, or inheriting one someone else controls.

The companies that win will not be the ones most dependent on a single media or activation ecosystem. They will be the ones that own their customer graph, control their intelligence layer, and preserve the flexibility to orchestrate across whatever partners, channels, and AI systems the future brings.

In the AI era, the customer graph is becoming the operating system for enterprise decisioning.

**

This thought leadership piece is written by Tony Owens, CEO, Amperity.