CES has long been a showcase for technological possibility. This year, we saw a noticeable market evolution, with less attention on features and tools and more focus on how organisations are restructuring themselves around AI as foundational infrastructure. Agencies, media owners, and leaders are starting to converge on a view we’ve long held: lasting value from AI depends on integrating models, data, workflows, governance and security into a coherent operating system.
This is a positive step forward. But significant challenges remain in operationalising AI at scale, particularly as AI agents proliferate across the industry. As brands transition from experimentation to full production deployment, infrastructure and hardware decisions beyond the cloud are critical to achieving true efficiency, control, and scalability. These foundational layers must also be recognised as essential components of the fully integrated AI stack that enables sustainable, long-term value.
From tools to systems
Early signals of this evolution in thinking are appearing across the agency landscape. At CES, leading digital-first media company Monks showcased the advancement of their agentic AI ecosystem – monks.flow – designed to unify planning, production and media operations into one efficient pipeline.
This is a great example of moving AI out of isolated use cases and into the core mechanics of how work is executed, recognising that without system-level integration, even the most advanced AI capabilities can accelerate fragmentation.
Where media and advertising is headed
More announcements at CES included NBCUniversal’s agentic approach to premium sports video buying that embeds AI directly into media execution. Omnicom and Stagwell showcased unified operating systems connecting strategy, creative, and activation through shared intelligence layers. Havas’ announcement of their AVA platform and WPP’s continued expansion of Open Pro further reinforced the direction of travel: proprietary platforms are becoming the primary interface between human decision-making and machine capability as opposed to disparate tools.
Why the full AI stack matters
What distinguishes these strategies is an understanding that AI outcomes are only as strong as the weakest layer in the stack. Access to models alone does not deliver efficiency or advantage. Without the integrated data environments, workflow orchestration, governance, and security, AI adoption risks increasing complexity rather than reducing it – not to mention the vital importance of infrastructure and hardware decisions that are often overlooked.
Effective AI integration requires:
- Coordinated model orchestration
- Controlled access to proprietary data
- Workflow-level embedding across teams
- Governance that scales with adoption
- Security designed into the system from the outset, including IP protection.
- Infrastructure and hardware decisions that optimise performance, cost, and control
When these layers operate together, AI becomes multiplicative, reducing cost and amplifying human judgement. When they do not, organisations see diminishing returns and escalating risk.
Who will capture lasting advantage?
CES 2026 revealed an important shift in mindset. The organisations positioned to capture lasting advantage are those treating AI as infrastructure – redesigning systems, workflows, and governance together rather than layering new capabilities onto legacy operating models.
As AI capability rapidly commoditises, architecture is becoming the differentiator. Those who invest in fully integrated AI stacks, from the ground up, will not only operate more efficiently but retain control over the intelligence that defines competitive advantage in an increasingly automated landscape.
At Stelia, this is the transition we’re focused on; helping organisations to architect AI systems that enable both capability advancement and IP control over the very assets that determine competitive positioning. Working with some of the major players in this space, we look forward to further progressing the AI operating systems on which some of the most innovative capabilities in the industry are emerging.
