Reframing AI for Real-World ROI: How IT Leaders Can Balance Innovation with Discipline

Introduction

It’s 2026. Your board is breathing down your neck about margins. Your CFO is eyeing every line of the technology budget with a microscope. Meanwhile, generative and agentic AI are flashing neon signs everywhere—promising transformational gains, but sometimes delivering little more than noise. CIOs and CTOs today face this dynamic: how to do more with less and invest smartly in innovation. 

In dozens of conversations with enterprise IT leaders across industries, one theme emerged consistently: cost savings remains the number-one priority. Macroeconomic uncertainty, margin compression, and shareholder expectations have turned IT investment into a sport of selective bets rather than full-scale leaps. Yet at the same time, leaders are under pressure to modernize operations, improve experiences (both customer and employee), secure their systems, and prepare for the coming wave of automation. 

The answer is not “either/or.” It’s “both/and.” This post is a blueprint for IT executives who want to avoid being dazzled by hype and instead architect a strategy that updates the fundamentals and creates innovation momentum. We'll walk through four key themes where I believe business-technology leaders should focus in 2026: 

1. Start with disciplined AI adoption — what I call Sequential Diversified Innovation (SDI).
2. Treat AI as the fourth channel — alongside voice, video, chat.
3. Embrace the Total Experience (TX) era — integrating CX + EX.
4. Elevate deep-fake resilience — because when trust breaks, everything else does.

1. Sequential Diversified Innovation (SDI): Innovation Without the Sideshow

AI is everywhere now. According to McKinsey & Company’s 2025 “State of AI” survey, more than three-quarters of organizations say they use AI in at least one business function. McKinsey & Company+1 Yet—and this is critical—over 80% of those organizations say they aren’t seeing material enterprise-level EBIT impact from AI today. McKinsey & Company+1

That disconnect is exactly why smart IT leaders need to apply discipline to their innovation agenda. I call this Sequential Diversified Innovation (SDI), and it has four steps:

Step 1: Start Small and Focused

Pick a use case where you can measure ROI within weeks or months—not years. Don’t chase the “moonshot” on day one. Maybe it’s automating post-call wrap-ups for your contact center, or deploying AI-driven spend analytics in procurement.

Step 2: Win Fast

Launch, measure, learn. Get a win under your belt. The value of quick execution isn’t just the modest cost savings—it’s credibility. Once the business sees real results, the next phase becomes much easier.

Step 3: Expand Sequentially

After your first win, roll out the next pilot—preferably in a different functional area (finance, HR, ops, supply chain). Diversify the investment so your AI portfolio isn’t a one-trick pony.

Step 4: Diversify Investment

This means don’t bet everything on one use case or one vendor. Portfolio diversification is as important in AI as it is in finance. 

Why this matters: Too many organizations overspend on AI in one silo—say a flashy chatbot—but ignore broader strategic areas. SDI shifts the mindset to sustainable value across functions. It’s not sexy, but it works.

2. Agentic AI as the Fourth Channel: Voice, Video, Chat …and AI  

Historically, customer-facing and employee interactions have cycled through three core channels: voice, video and chat. In 2026, we need to recognize a fourth channel: AI. 

Think of it like this: if voice, video and chat were the band, then AI is the drummer changing the rhythm. It sits beside your employees, not in front of them: 

• It assists by providing real-time insights.
• It automates routine workflows.
• It enables smarter interactions across every channel. 

Start with some data: today roughly 20% of interactions are person-to-person (P2P), about 60% are person-to-machine (P2M), and 20% are machine-to-machine (M2M). By 2030, M2M interactions may climb above 22% as autonomous agents start doing the buying, planning, and servicing. 

In this context, AI is no longer a feature—it’s a full-blown interaction channel. If your voice/video/chat stack were the band, AI is adding a whole new section of instruments. 

What this means for CIOs/CTOs: 

• Build your architecture so AI can plug in across channels, not just on top of them.
• Governance: ensure AI-enabled interactions are aligned with identity & access policies. This ties directly to deep-fake risk (more on that later).
• Channel strategy: don’t treat AI as a “bot” channel only. Instead, enable it to enhance voice, video, and chat. 

3. The Era of Total Experience (TX): CX + EX Converge

2026 isn’t just about customer experience (CX). It’s about Total Experience (TX)—the combined sum of CX and employee experience (EX). Why? Because every internal inefficiency, every clunky tool used by an employee, bleeds into your CX. Employees suffer, customers eventually notice. 

My formula: 

TX = Sum of CX UX scores + Sum of EX UX scores 

And here’s the kicker: according to Forrester Research, firms who lead in TX enjoy 1.6× higher customer satisfaction and 1.9× higher employee engagement than their peers. 

To win in TX: 

• Evaluate the internal “agent experience” with the same rigor you evaluate external CX.
• Invest in enterprise UX, not just customer-facing apps.
• Make the data flows seamless—from employee tools to customer channels to analytics platforms.

4. Deep Fakes: When Authenticity Becomes an Enterprise Risk

Here’s where strategy meets risk. We often talk about AI in terms of productivity. But we must also confront the darker side: deep fakes. These are AI-generated audio, video, or images designed to mimic real people—and they’re becoming dangerously good. 

Deep fakes aren't just fun viral clips anymore. They’re enterprise threats: executive impersonation calls, synthetic voices authorizing transactions, manipulated video messages sent to customers or partners. The target? Trust.

Key investment focus for 2026:

• AI-Powered Detection & Verification: Deploy models that detect synthetic media before it hits your organization’s communication channels.
• Digital Identity Reinforcement: Strong, behavior-based identity and access controls that don’t rely purely on voice or face recognition (both can be spoofed).
• Governance & Awareness: Treat manipulated media as a core incident category. Train both executives and frontline staff to recognize and escalate.
• Provenance & Watermarking: Establish trusted pipelines and metadata for audio/video content. If you can’t trace it, you can’t trust it.

5. Putting It All Together: A CIO’s Four-Point 2026 Agenda 

Here’s your playbook for 2026, distilled: 

1. Adopt AI with discipline (SDI): start small, win fast, expand sequentially, diversify investment. 
2. Treat AI as the fourth channel: plug AI into your voice/video/chat ecosystem.
3. Lead with Total Experience: evaluate UX internally and externally.
4. Defend authenticity: deep-fake detection and trust assurance are non-negotiables. 

6. Example of What it Looks Like

Here’s a fictionalized but representative scenario to illustrate the strategy: 

A mid-tier financial services organization had already deployed chatbots in their customer service area—but hadn’t seen meaningful gains. The CIO pivoted: she selected a wrap-up automation use case in the contact center (small, focused). They measured time-spent per call, redeployed that freed time to agents, and rolled it out to operations. Next, they plugged in an AI-driven ‘recommendation engine’ into the voice channel (AI as the fourth channel). At the same time, the infrastructure team redesigned agent desktops to remove redundant tools (TX). Finally, the security team embedded deep-fake detection into executive comms and voice authorization flows (authenticity). The result? ~12 % reduction in cost per interaction, improved agent satisfaction, and zero major impersonation incidents all year. 

That’s the SDI + fourth-channel + TX + authenticity stack in action. 

7. Three Leadership Imperatives for CIOs 

1. 1. Lead with structure, not just technology. According to McKinsey’s report, redesigning workflows had the biggest effect on bottom-line for AI deployments. McKinsey & Company+1
   2. Embed governance early. The same survey shows only ~28% of AI-using firms have a CEO responsible for AI governance. That needs to change. Board Agenda
   3. Treat trust as infrastructure. Your digital trust architecture should be as robust as your network or identity stack. Deep fakes don’t care about perimeter firewalls.

Conclusion

2026 isn’t the year to throw full weight into a flashy AI project that lacks business alignment or governance. It’s the year to build an intelligent, resilient, trustworthy foundation—one where innovation happens within the architecture of discipline and value. 

If you can win small, connect AI to your core channels, orchestrate all experience touchpoints, and protect authenticity at scale—you’re not simply keeping up. You’re leading. 

Let’s make next year the year where AI isn’t just hype. It becomes business muscle.