Guardrails, Privacy, and the Future of Applied AI

AI is moving fast. Too fast, some would argue. In telecom infrastructure, speed must always be matched with safeguards. That was one of the most important messages to come out of the Connect(X) 2025 Spotlight Panel, hosted by the Wireless Infrastructure Association. 

The discussion featuring Inorsa CEO Sean Shahini and Harmoni Towers VP D.J. Grosso made one thing clear: AI isn’t just about automation. It’s about trust. Without privacy and guardrails, the risks outweigh the benefits. 

The Privacy Question 

For carriers, towercos, and vendors, data is gold — leases, RFDS packages, structural drawings, and compliance reports all contain proprietary information. Losing control of that data isn’t just a financial risk. It’s a reputational and regulatory one.  

D.J. Grosso put it simply: 

“Data privacy is everything to us. Contracts are proprietary. They have to stay within our organization.” 

Sean emphasized that outsourcing work to 60–70% overseas firms — a common industry practice — is both inefficient and risky. “That’s not just slow. It’s unsafe,” he warned. 

The Inorsa Approach: Private AI 

So what’s the answer? Inorsa’s solution is Private AI: customer-specific models that never mix data across organizations. Each client gets its own AI engine — effectively a digital employee that never forgets, never shares, and always works within your walls. 

As Inorsa CTO Zain Admani explained in a separate conversation: 

“There really is one major concern: are you training on my data? Is everyone else going to have access to it? We self-host our models so we can guarantee data never leaves and never cross-trains across customers.” 

That approach delivers two things customers demand most: control and confidence.  

Guardrails That Matter 

AI without oversight creates more problems than it solves. That’s why Inorsa builds compliance guardrails directly into workflows: 

• Automated field validation ensures errors and conflicts are flagged before they spread downstream. 
• Role-based access keeps sensitive data visible only to the right stakeholders. 
• Self-hosted models ensure data is never exposed to third parties. 

Sean’s view is unapologetic: 

“Our role is to make customers the best in class, so they can scale faster and use their teams to build a bigger business.” 

That means giving customers automation they can trust — automation that moves fast without breaking things.  

Why Guardrails Create Speed 

It may sound counterintuitive, but guardrails don’t slow things down. They enable speed by preventing rework, ensuring compliance, and reducing risk. 

As Zain explained: 

“If your data is being mixed with others, you can’t guarantee safety. By self-hosting, we remove that risk and give customers confidence. That confidence is what accelerates adoption.”  

Confidence is critical. Without it, teams hesitate to automate. With it, they lean in — and the compounding gains of efficiency, collaboration, and speed to revenue kick in. 

Why Now 

The clock is ticking. With FCC AI policy updates shaping compliance and spectrum expansion creating pressure across the industry, waiting isn’t an option. 

Sean summed it up at Connect(X): 

“This train is already leaving. In the next 12 months, most services will be highly automated. Adaptation becomes harder the longer you wait.” 

The message for telecom leaders is clear: adopting AI with guardrails isn’t just smart. It’s survival. 

The Takeaway 

Applied AI isn’t about hype or fear. It’s about speed, trust, and control. The first AI-enabled service platform for infrastructure assets gives telecom leaders exactly that — automation with the privacy and safeguards they demand.  

The future belongs to the companies that move fast with confidence. Inorsa is building that future. 

👉 Ready to see how? Get a demo today. 

👉 Or watch the full Connect(X) panel recording for the complete discussion.