Second in a four-part series unpacking the pillars of CASA26.

A year ago, the story of AI in the Middle East was a story of pure ascent. The Gulf had decided it would not merely consume artificial intelligence; it would build the ground it runs on. Abu Dhabi broke ground on a gigawatt-scale compute cluster. Saudi Arabia’s sovereign wealth stood up a national AI champion and raised its ambitions into the trillions. Washington cleared the export of tens of thousands of advanced GPUs that had been frozen for months. With more than three trillion dollars of sovereign capital behind them and regional data-centre capacity set to triple by the end of the decade, the Gulf states could build while everyone else waited for financing.

Then, in late February, the other face of the bet arrived. As the conflict with Iran escalated, drone and missile strikes hit data-centre infrastructure across the Gulf — and AWS confirmed its UAE region had been knocked offline, with recovery measured in months and customers told to move their workloads elsewhere. The region that had marketed itself as a safe harbour for the world’s data watched a flagship cloud region burn. In the space of twelve months, sovereignty went from a slide about data residency to a question with missiles attached: who physically controls the ground intelligence runs on — and what happens when that ground is contested?

That is the real subject of CASA26’s second pillar — AI, sovereign infrastructure and agentic systems. It’s also why e& enterprise’s Ahmed Omer is returning to CASA26 to talk through sovereignty from a Middle East vantage point that looks nothing like it did a year ago.

The hard part is no longer building the model

For most of the AI cycle, the centre of gravity sat in a handful of hyperscaler data centres where the largest models were trained. That era is closing. Models are commoditising — capable, open-weight, increasingly interchangeable. The defensible value is moving to a different question.

The hard part is no longer building the model. The hard part is deciding where it runs.

Training is a one-time, centralised act. Inference — the actual work of intelligence, performed millions of times a day — is not. It wants to be close to the data, close to the user, close to the decision. That pulls it out of the hyperscaler core and toward the edge: into networks, devices, enterprise premises and regional infrastructure. The economics point the same way. The binding constraint on AI is no longer chips alone; it is power, latency and the cost of moving data around. Inference at the edge is often cheaper, faster and easier to govern than inference shipped to a distant cloud and back.

This is the opening the telecom industry has been waiting for, whether or not it realises it.

Why this is the operators’ second chance

Hyperscalers own scale. What they cannot easily own is proximity, regulatory standing and trust — the three things that decide where regulated intelligence is allowed to run. Operators have all three almost by default: physical infrastructure close to the user, deep compliance relationships with national regulators, and decades of being trusted with sensitive data.

Sovereign AI is usually framed as a burden — GDPR, the EU AI Act, data-residency rules, the long list of reasons a regulated enterprise cannot simply pipe its data into a centralised model. Framed that way, it’s a cost. Framed correctly, it’s a market. The very rules that make centralised LLMs unusable for a bank, a hospital or a government create demand for intelligence that runs on controlled, in-country, compliant infrastructure. Someone has to host that. The operator with edge sites, a power footprint and a regulator’s trust is better placed to do it than almost anyone.

We saw the early shape of this at CASA25. Deutsche Telekom set out a blueprint for a sovereign CPaaS built on European terms. KPN made the case that reinvention is as much about AI culture and capability inside the operator as about the technology itself. Intel joined the uncomfortable conversation about why telcos keep failing at innovation — which is, at heart, this exact gap between holding the right assets and actually moving on them. And BT’s work on UC Edge points at the same instinct: push capability out to where the customer and the data already are.

The assets are real. The question CASA26 has to answer is whether the industry can convert them into a commercial position before the window closes.

Agents make the question urgent

There is a third force tightening all of this: agentic AI. Once intelligence is distributed across networks, devices and enterprises, the next step is agents that don’t just answer but act — booking, paying, negotiating, resolving, on a customer’s or a company’s behalf, with a degree of autonomy.

That changes the stakes of “where it runs” entirely. An agent acting autonomously inside a regulated business is not a chatbot; it’s an actor with access, authority and consequences. The control plane — the layer that governs which agents may operate, on what data, under whose rules, with what audit trail — becomes the most valuable real estate in the stack. It is, in effect, the new orchestration layer, and it sits naturally next to the infrastructure that hosts the inference.

CASA has to be honest here, because the industry often isn’t. Much of what is sold today as “agentic AI” is still a demo. At CASA25, the agentic sessions — including the “beyond the buzz” discussion Kyle Nel led, with Intel among those at the table — kept returning to the unglamorous questions: consent, identity, trust, governance, and whether any of it survives an enterprise procurement process. The capability is racing ahead. The governance, the economics and the trust model are not. That gap is the opportunity, not the disappointment.

What CASA26 will do with this pillar

The AI and sovereignty track at CASA26 is built around the questions that decide who captures this — not the ones that demo well:

• Where will inference actually run in five years, and who pays for the power, the edge sites and the compliance?
• Is sovereign AI a genuine commercial line for operators and regional infrastructure players, or a slogan they’ll cede to the hyperscalers anyway?
• What does the control plane for agentic AI look like in practice, and who is positioned to own it?
• And from the Middle East to Europe, how does a region turn geopolitical exposure into a reason to keep intelligence on home soil?

These are analyst-led discussions, in a room of around 150 senior leaders who build, regulate, finance and deploy this infrastructure — operators, chipmakers, investors and the regional voices, including e& enterprise, who are living the sovereignty question in real time.

This is also one of the CASA26 tracks open for a partner to help shape and lead. If your company’s future depends on where intelligence runs — and on being seen as one of the organisations defining the answer — this is the conversation to be inside, while the track is still open.

The cloud was built for humans. The next infrastructure is being built for intelligence — and the question of whose soil it stands on is no longer academic.

Amsterdam. September 21–23. The next chapter starts now.

---

Next in the series: Network APIs & Telecom Transformation — from API counts to programmable networks, identity, and the business model operators keep missing.

At CASA25, amid the intense discussions on network APIs, GTM models, and CPaaS monetization, one panel stood out for its mix of raw honesty, technical depth, and forward-looking pragmatism: Agentic AI — Consent, Identity & Trust.

Moderated with humor and curiosity by Kyle Nel of Sandbox Industries, this panel cut through the AI hype and tackled the hard questions: What really makes AI “agentic”? How do enterprises get started without getting lost? And how do we prepare humans — not just machines — for the agentic era?

Here are the highlights and takeaways from the session featuring:

• Kyle Nel (Sandbox) – Moderator, venture investor, AI outsider-turned-ally
• Adnan Saleem (Radisys) – CTO of Software & Cloud Solutions
• Mark Castleman (Intel) – Ecosystem lead for AI and deep tech
• Olaf Wallaart (CM.com) – Lead Data Scientist, Agentic AI platform
• Remco Magielse (GLBNXT) – Startup founder tackling the “proof of concept to scale” gap

🔁 From AI Hype to Agentic Reality

Kyle opened the discussion with a simple question: “Who here is actually using agentic AI?” A few hands went up — but more had used AI in their personal lives. As he noted, the consumer world is once again leading the enterprise, signaling what’s to come.

But as Olaf Wallaart pointed out, most organizations are still trying to figure out the basics — how to write a proper prompt. The magic agent that handles everything on its own? “It doesn’t exist,” said Castleman. “You need a hundred agents, each tuned to a specific outcome, with checkpoints along the way. It’s a recipe — not a single ingredient.”

Agentic AI isn’t just a smarter chatbot. It’s the orchestration of intelligent agents capable of executing tasks, calling APIs, and interacting with other systems. But the infrastructure of most businesses — especially telcos — is still too fragmented, siloed, and legacy-heavy to support this at scale.

🎯 Where to Start: Low Complexity, High Impact

Remco’s team uses a simple framework to assess use cases: map complexity vs. impact. The goal? Find the “wow” moments — high-impact wins that are easy to implement and explain on a single slide.

“You don’t want safe,” he said. “You want something that gets people to go: Whoa, this changes how we work.”

Adnan echoed the point: “Don’t aim for sci-fi agents yet. Start with low-hanging fruit that improves productivity, then build from there.” Internal tools — like secure ChatGPT-style assistants — are often the best starting point.

🧠 AI Literacy Is the New Digital Literacy

The panel agreed that a huge part of the challenge is human, not technical. We’ve spent decades teaching people to write short keyword queries. Now, we must teach them to write long, rich prompts — to treat machines like collaborators, not search engines.

Olaf shared how CM.com formed an internal AI team tasked with helping every department identify their top problems — then co-create solutions. “They become like a beacon of AI, promoting ideas, showing what’s possible, and building internal momentum.”

Training helps. Practice is better. And curiosity is the ultimate success signal.

“Curious people engage deeply,” said Castleman. “Transactional ones don’t. Look across your org — who’s asking better questions? Those are your early adopters.”

⚙️ Infrastructure, ROI, and the Spaghetti Stack

Agentic AI reveals the cracks in enterprise systems. Legacy platforms, dirty data, and siloed infrastructure become instantly visible when you try to automate tasks end-to-end.

🤝 Telcos: From Dumb Pipes to Agentic Enablers

The final part of the panel tackled the telco opportunity. As Castleman put it bluntly: “In the old app era, telcos lost because they tried to figure out the apps instead of enabling them. This time, they need to expose all the little levers and let the agents go wild.”

Agentic AI breaks apps down into atomic functions. Telcos and CPaaS players sit on a goldmine of these functions — but they must make them available as APIs that agents can trigger.

Identity, trust, and consent become key. As Olaf noted, telco networks have something unique: reliable identity. That’s a massive asset in an era where bots can imitate humans — and vice versa.

🧪 Final Words: It’s About Agency — Not Just AI

The last round of the panel shifted from tech to philosophy.

“If done right, AI gives us back our agency,” said Olaf. “We get to build quickly without begging for budget. We get to focus on what’s fun and impactful.”

AI might reduce cost. But the panel’s message was clear: it’s about productivity, creativity, and revenue growth — not just optimization.

And it’s about people. Give them tools. Train them. Let the curious ones lead.

Because in the end, agentic AI isn’t just software. It’s a shift in how we work, build, and think.

Insights from Gamma, Sandbox Industries, and Intel at CASA25

When you put a telco CEO, a venture capitalist, and a technology investor on the same stage and ask them to “fix telco innovation,” you don’t expect consensus.

At CASA25, Nicolai Schaettgen (Match-Maker Ventures) did just that—casting Mike Mills (Gamma) as the telco CEO, Dan Phillips (Sandbox Industries) as the corporate shareholder, and Mark Castleman (Intel) as the tech and investment veteran.

The result was an unfiltered reality check on why innovation in communications keeps stalling—and what needs to change.

1. The innovation illusion

Telcos, Mills admitted, know where they want to go—but can’t always find how to get there.

“We know our strategy. We know the direction. What’s missing is finding capabilities that aren’t mainstream—investing where one plus one can become three.”

For years, telcos have poured money into “innovation” through corporate venture capital (CVC) arms, incubators, and R&D labs. But the hit rate is abysmal.

Why? Because most telcos still apply improvement logic to innovation problems.

2. You can’t get ice cream from a meat grinder

Castleman, who’s raised hundreds of millions as a founder and investor, put it bluntly:

“Large systems like telcos are designed to improve what they do, not to deviate. Every output is the perfect output of that system. If you want something different, you need a different system.”

Innovation, he argued, means deviation, not optimization. You can’t expect a compliance-driven, quarterly-report-focused machine to suddenly start taking moonshots.

“You can’t put ice cream ingredients into a meat grinder and expect ice cream,” he joked. “You need a different system.”

The reason most corporate venture units fail is because they’re still tethered to the parent company’s antibodies—its incentives, KPIs, and politics.

3. Why telco CVCs don’t work

When Schaettgen asked who would still set up a corporate venture fund, none of the panelists raised a hand.

Castleman:

“If I’m not in the business of making bets, how am I suddenly going to succeed at venture investing? Better to put that capital into professional funds that already have access, insight, and deal flow.”

In other words, invest in the investors.

Don’t pretend to be one.

For telcos, the goal isn’t financial return—it’s access to innovation.

That access can be achieved far more effectively by becoming a limited partner (LP) in specialized funds or joining curated ecosystems—like the CPaaS Acceleration Alliance—where visibility, not ownership, creates advantage.

4. What shareholders actually want

Phillips, representing the shareholder view, described the telco dilemma succinctly:

“My focus is growth and margin. My frustration is that telcos build the infrastructure, but others capture the 90 percent gross margins and giant exits.”

To many investors, telcos look like annuities—slow-growth, dividend-paying utilities.

When those companies start putting cash into “high-risk, low-promise” innovation bets, shareholders see it as their dividend being spent.

The smarter play?

Spread smaller investments across a wider ecosystem. Gain visibility into innovation rather than trying to control it.

“What you really want is clarity so you can reduce your five-year risk horizon,” Castleman said. “Spread visibility of access—not ownership.”

5. What it takes to absorb innovation

Even if telcos partner with external innovators, success depends on internal readiness.

Phillips warned that without structure and leadership commitment, innovation dies in the middle:

“There needs to be organizational buy-in from the CEO down, and a process for working with startups—or scaleups. Otherwise, things stall after a few pilots.”

That means:

Dedicated processes for contracting and onboarding smaller partners (not the same RFP that takes three years for a core system). Clear KPIs for pilots—defined outcomes, time limits, and ownership. Different language—talk about scaleups and tech partners, not startups. The word alone is a red flag in boardrooms.

Mills agreed:

“The clash between telcos’ need for scale and compliance and a startup’s need for speed is very real. Managing that cultural challenge is hard—but necessary.”

6. Startups: grow up

The other half of the problem lies with the innovators themselves.

Castleman reminded founders that working with a telco isn’t a playground—it’s mission-critical infrastructure.

“If your network goes down for minutes, you can calculate churn by the minute. Startups need to stop acting like startups and start acting like grown-ups.”

Innovation doesn’t mean chaos. The goal isn’t to stay a startup forever—it’s to mature fast enough to meet enterprise-grade standards.

As he quipped:

“A six-year-old startup is a startup that’s been failing for four years.”

7. The CPaaS angle: sell outcomes, not features

Finally, the conversation turned to the CPaaS perspective.

Castleman reframed how platform players should approach telcos:

“If I’m a CPaaS player, I’m going to go to the telco CEO and say: I understand your outcome. I’ll help you achieve that with X percent more growth. Let’s define it together and hold me accountable.”

This is the opposite of a typical tech pitch—no laundry list of APIs, no “400 tools to help you do more.”

It’s co-creation around measurable outcomes.

And it’s exactly how intelligent engagement and network-API ecosystems need to evolve.

8. The CASA25 takeaway: change the system

Schaettgen closed the session by thanking the panel for the candid discussion—and for playing along with his “role-play therapy for telcos.”

The verdict was unanimous:

Stop building CVCs that mimic VCs but operate like procurement departments. Invest in visibility through specialized funds and innovation alliances. Build internal processes that can actually absorb external innovation. Speak the right language—scaleups, not startups; outcomes, not ideas. Remember: innovation isn’t about tweaking the machine. It’s about building a new one.

Epilogue: From CASA25 to CASA26

The conversation at CASA25 echoed a growing truth across the industry: the future belongs to ecosystems that can learn faster than any single company can innovate alone.

As capital, cloud, and communications continue to converge, the winners will be those who stop guarding the gate—and start opening it.