AI adoption is accelerating across every industry—but for enterprise leaders, the real question isn’t whether to use AI. It’s where and how AI should run safely and effectively, especially on mobile devices where personal and professional data converge.

In a recent Stratix podcast episode, Alex Kalish, Chief Strategy and Solutions Officer at Stratix, sat down with Ian Marsanyi, Google Pixel for Business, and Sean Gideon, Head of Partnerships, Android Enterprise, to explore what it takes to operationalize AI at the edge—without compromising governance, security, or productivity.

Here are the key takeaways for organizations navigating this rapidly evolving landscape.

What does “AI at the edge” mean for enterprises?

AI at the edge refers to running AI workloads directly on devices like smartphones, tablets, and other endpoints—not just in the cloud.

For enterprises, this matters because mobile devices sit at the center of daily operations. They capture sensitive data, support frontline workflows, and operate in environments where connectivity isn’t always guaranteed.

The edge is where work actually happens. That makes it one of the most important—and most complex—places to deploy AI.

Should AI run on-device or in the cloud?

The short answer: both.

The reality is that AI in the enterprise is becoming hybrid by design. Some workloads are best handled locally on the device, while others require the scale and intelligence of the cloud.

On-device AI tends to perform best when speed, privacy, or connectivity are critical. Think of frontline workers who need immediate responses, or use cases where data should never leave the device.

Cloud AI is better suited for complex analysis, aggregating data from multiple systems, or generating insights that require broader context.

The key question isn’t where AI should run in general—it’s where it should run for a specific use case. That decision should be guided by the data involved, the workflow requirements, and the expected outcome.

Why is governance such a challenge with AI on mobile devices?

Governance is struggling to keep pace with how quickly AI is being adopted.

Many organizations are putting policies in place, but at the same time, encouraging teams to move quickly with AI. That tension creates gaps—especially at the user level.

One of the biggest risks emerging is “shadow AI.” Employees, trying to be more productive, often turn to consumer-grade AI tools that haven’t been approved by IT. They’re not trying to create risk—they’re trying to get their jobs done faster.

But those tools can expose sensitive data, create compliance issues, and operate outside any governance framework.

The underlying issue isn’t just policy—it’s usability. If approved tools aren’t accessible or effective, users will find alternatives.

How should enterprises rethink AI governance?

A critical shift is moving from feature-based governance to data-based governance.

Instead of trying to control individual AI features, organizations need to focus on what data is being accessed, how it’s being used, and where it’s flowing.

This approach is more durable. AI capabilities will continue to evolve rapidly, but data governance principles remain consistent.

It also creates a more scalable model. Rather than reacting to every new feature, enterprises can define rules around data sensitivity and apply those rules across all tools and platforms.

Why is context so difficult for AI on mobile devices?

Context is one of the hardest problems in AI—especially on mobile.

The same interaction can mean completely different things depending on the situation. A query that looks personal in one moment could be a work-related request in another.

Mobile devices make this more complex because they are inherently personal. They hold a mix of personal and business data, and they’re used across different environments throughout the day.

AI systems need to understand not just the query, but the intent, the user, and the environment in which it’s happening. Without that context, it becomes difficult to apply the right governance policies or deliver the right outcomes.

How is AI changing enterprise device strategy?

AI is fundamentally changing what it means for a device to be “enterprise-ready.”

In the past, mobile devices were essentially access points—for email, messaging, or basic applications. Today, they are becoming active compute platforms capable of running AI workloads directly.

This shift means device decisions now have long-term implications. The hardware an organization chooses today will influence what AI capabilities it can support over the next several years.

Performance, memory, and processing power are no longer just technical specs—they are enablers of future innovation.

Organizations that invest in higher-capability devices are better positioned to support advanced AI use cases, maintain performance, and reduce dependency on constant cloud connectivity.

How does device strategy impact AI governance?

There’s a direct relationship between device capability and governance risk.

When devices can’t support secure, efficient AI workflows, users often look for alternatives. That’s where shadow AI begins to emerge.

In this way, procurement decisions can unintentionally shape governance outcomes. Choosing devices without sufficient capability can push work into less controlled environments.

A strong device strategy helps close that gap—ensuring employees can access the tools they need without compromising data security.

What role does voice and speech-to-text play in AI adoption?

Voice is becoming one of the most important—and most underutilized—interfaces for AI.

For frontline workers, typing into a device isn’t always practical. Voice enables a more natural and efficient way to interact with systems, capture information, and complete tasks.

Speech-to-text capabilities, in particular, can dramatically improve data quality and productivity. Instead of minimal or incomplete inputs, organizations can capture richer, more usable information from the field.

As AI continues to evolve, voice will play a central role in how users engage with technology across environments.

How should enterprises approach AI moving forward?

AI at the edge requires a balanced approach. Organizations need to enable innovation while maintaining control—but those goals don’t have to conflict.

Success comes from aligning AI initiatives with real-world workflows, designing governance around data rather than features, and ensuring employees have access to tools that are both powerful and practical.

It also requires a shift in mindset. AI isn’t just another layer of software—it’s fundamentally changing how work gets done.

Enterprises that embrace that shift thoughtfully will be better positioned to drive productivity, protect sensitive data, and unlock the full value of AI across their operations.

Final Thoughts

AI at the edge represents a powerful opportunity—but also a complex balancing act.

Organizations that succeed will be those that:

• Align AI with real-world workflows
• Design governance that adapts with innovation
• Empower employees without compromising control

Because in the end, AI isn’t just about technology—it’s about how people get work done, wherever that work happens. Want to dig deeper into the topic? Click here to listen to the podcast.