CES 2026 opened yesterday with keynotes, panel discussions, and enterprise gatherings showcasing breakthrough innovation across industries. Yet at the same time, the conference began by spotlighting a critical paradox: while AI capabilities are advancing faster than most organisations can absorb them, the path to capturing lasting value remains constrained by short-term thinking and weak architectural foundations.
AI has moved decisively from experimentation to production deployment, and across the event’s opening activities – from C Space’s panel on AI’s creative evolution to Fortune’s Brainstorm Dinner examining agentic AI in enterprise operations – the conversations all pointed toward one critical differentiator: the organisations succeeding at scale are those prioritising architectural foundations over capability adoption.
Below, we examine the key signals from the first day of CES 2026 and what they reveal about the architectural imperatives organisations must prioritise to move beyond AI experimentation.
The creative evolution: strategic potential constrained by short-term thinking
Monks and Leonardo.ai’s C Space panel discussion on the creative evolution of AI kicked off the event, moving the conversation of AI impact beyond efficiency gains to explore how technology enables strategic organisational change.
The panel’s insights, featuring Sir Martin Sorrell (S4 Capital), Sarah Jost (Amazon Ads), and Dwayne Koh (Leonardo.ai), highlighted the shift in thinking necessary to capture AI’s full potential. While AI is removing traditional barriers in media production and democratising storytelling at scale – enabling anyone to become a high-level storyteller and allowing brands to deliver precision storytelling across audiences – the path to lasting organisational impact remains constrained.
Sorrell emphasised that AI’s greatest potential lies in destroying information silos and flattening organisational structures. Yet heavy adoption, he noted, remains driven primarily by external economic pressures rather than strategic vision. All of this culminates in companies remaining too focused on short-term execution and sales to fully embrace AI’s strategic capabilities.
This tension, between AI’s transformative potential and organisations’ tactical deployment approach, set the tone for discussions throughout the day. And the answer is becoming clear: for AI capabilities to deliver sustained competitive advantage, organisations must invest now in the purpose-built systems that enable resilience and adaptability at scale, taking a disciplined, problem-first approach to adoption rather than racing for early, unsustainable wins.
Enterprise realities: agentic AI and humans in the loop
Indeed, this challenge isn’t isolated to one panel discussion, nor the creative industry alone. Ula Nairne, Stelia’s VP Media & Entertainment, joined senior executives across industries convening at Fortune’s Brainstorm dinner who shared similar concerns about the complexities of deploying autonomous AI systems at scale.
The evening focused on the realities of agentic AI within enterprises, examining how organisations are navigating the transition from AI as a support tool to AI as an autonomous operational layer. Crucially, the “Humans in the Loop” discussion – featuring leaders from Deloitte, The Walt Disney Company, Salesforce, and Solventum – underscored a critical tension emerging across enterprise deployments: as AI systems assume greater autonomy, questions of governance, oversight, and architectural resilience become operational imperatives. Long-term tangible impact through AI initiatives now hinges on purpose-built systems with governance embedded from the ground up.
These conversations reinforce the challenges around production-scale deployment that our partners are increasingly bringing to us. Decidedly, organisations moving beyond proof-of-concept into production deployment are confronting challenges that have less to do with model capability and more to do with the infrastructure required to effectively support reliable, secure, and economically viable AI systems at scale.
Infrastructure announcements: meeting enterprise AI demands
While enterprise leaders debated deployment strategies, hardware announcements from NVIDIA, AMD, and Intel underscored the infrastructure required to support AI at production level.
NVIDIA unveiled its Vera Rubin platform, promising 5x greater inference performance and roughly one-tenth the cost per token, addressing economic pressures across enterprise. AMD CEO Lisa Su announced the company’s MI455X accelerator, claiming 10x performance increases for AI workloads, alongside new Ryzen AI 400 series processors and Helios rack-scale systems designed to compete in the enterprise AI market. And Intel launched its Core Ultra Series 3 processors, the first to use the company’s 18A chip technology.
These announcements reflect a market directly responding to the demands of enterprise-grade AI deployment. Yet hardware advances alone don’t guarantee successful deployment. These advances represent pieces of a broader architectural puzzle, one that requires viewing compute, orchestration, and governance as integrated systems rather than standalone components in order to function cohesively at scale.
These developments reinforce the same consistent theme: AI capability advancement is outpacing organisational readiness. While hardware providers are scaling infrastructure to meet production demands, it will be the architectural decisions organisations make today about how to deploy, govern, and secure these systems that will ultimately determine which are able to capture lasting value.
The emerging pattern
The team’s first day at CES uncovered a consistent narrative across panel discussions, enterprise gatherings, and infrastructure announcements: while AI is transitioning decisively from experimentation to production deployment, the path to capturing lasting value hinges on how organisations choose to architect their systems.
The organisations succeeding at scale will be those that recognise AI deployment as fundamentally an architectural task, embedding flexibility and resilience from the start rather than retrofitting capabilities as requirements evolve, and treating governance not as a constraint but as the foundation that enables innovation to scale.
As discussions and developments continue throughout CES 2026 this week, the Stelia team remains focused on these architectural imperatives – partnering with organisations navigating the transition from AI experimentation to production infrastructure and facilitating sustained competitive advantage.
