Arm Holdings is positioning itself at the heart of the next major shift in artificial intelligence. In a recent podcast interview, Vince Jesaitis, Arm’s head of global government affairs, shared how the company sees AI evolving, why edge computing matters, and what this transition could mean for enterprises, governments, and the wider technology ecosystem.
Arm Editorial Team
From the cloud to the edge
For the past decade, AI progress has largely been associated with cloud computing and vast data centres. Training large models and running inference remotely has driven many of today’s AI services. Arm believes that model is starting to change.
According to Jesaitis, the next phase of AI will be defined by decentralisation. Instead of relying solely on distant cloud servers, more AI workloads, particularly inference, are expected to run locally on devices. This includes everything from smartphones, earbuds, and cars to industrial machinery and sensors.
“The next ‘aha’ moment in AI is when local AI processing is being done on devices you couldn’t have imagined before,” Jesaitis said.
Arm is well placed for that shift. Over the past year alone, its intellectual property has powered more than 30 billion chips worldwide, embedded in everyday consumer electronics and critical infrastructure alike.
Why edge AI matters
Arm highlights three main advantages of moving AI closer to where data is generated.
First is efficiency. Arm-based chips are designed for low power consumption, which helps reduce energy use and cooling requirements. That translates into lower operating costs and a smaller environmental footprint, an increasingly important factor for organisations under pressure to meet sustainability targets.
Second is speed. When AI models run locally, latency drops dramatically because data no longer needs to travel back and forth to the cloud. This is crucial for applications such as real-time translation, responsive industrial control systems, and instant safety triggers in industrial IoT environments.
Third is security and privacy. Keeping data on-device limits the need to send sensitive information off-premise. For regulated sectors such as healthcare, finance, and manufacturing, this can simplify compliance. Even organisations handling less sensitive data are paying closer attention to this approach as cyber threats and data breaches continue to rise.
Taken together, these benefits point to a future where AI is embedded throughout physical environments, rather than concentrated in a handful of massive data centres.
Working with governments worldwide
Arm’s strategy goes beyond technology. The company is deeply engaged with policymakers around the world, particularly as governments compete to strengthen domestic semiconductor capabilities following the supply chain disruptions seen during the COVID-19 pandemic.
Jesaitis said Arm is actively involved in workforce development, including collaboration with the White House on education initiatives aimed at building an AI-ready workforce. From Arm’s perspective, technological independence depends as much on skilled people as it does on access to hardware.
Regulation is another area where Arm seeks balance. The US tends to prioritise innovation and speed, while the European Union focuses more on safety, privacy, and legally enforced standards. Arm aims to operate between these approaches, designing technologies that meet strict global requirements while still supporting rapid AI advancement.
The enterprise case for edge AI
For enterprises, Arm argues that edge AI offers a compelling mix of scalability, security, and cost control. By avoiding full centralisation in the cloud, organisations can deploy AI where it is needed without exposing themselves to certain risks associated with shared infrastructure.
Arm is also investing heavily in hardware-level security, which can help mitigate issues such as memory exploits that are often beyond the control of end users relying on centralised AI services.
Regulatory pressure is unlikely to ease in any sector. In fact, most industries can expect tighter rules and higher penalties for non-compliance. In this environment, companies that can demonstrate secure, efficient systems may gain a competitive edge. Arm sees edge AI as a natural fit for that future.
Sustainability is another factor, particularly in Europe and Scandinavia, where ESG goals are becoming central to technology decisions. The energy-efficient design of Arm-based chips aligns with this trend. Even major cloud providers are responding, as seen in Amazon Web Services’ growing use of Arm-based platforms for low-cost, low-power workloads.
What comes next
Looking ahead 12 to 18 months, Jesaitis expects continued growth in global AI capacity, driven by exports and large providers in regions such as the US and the Middle East. Arm plans to support both hyperscalers and the expanding demand for edge-based AI across industries.
He also sees edge AI playing a key role in addressing concerns about the environmental impact of computing. With a long history in mobile and low-power devices, Arm’s technology is inherently suited to organisations trying to balance performance with energy efficiency.
Redefining what “smart” means
Arm’s vision of edge AI is about more than convenience. It is about systems that are context-aware, responsive, secure by design, and affordable to operate. With near-zero latency and local intelligence, devices no longer need to rely on constant connectivity to be effective.
As Jesaitis put it, “We used to call things ‘smart’ because they were online. Now, they’re going to be truly intelligent.”
