What To Do About AI in 2026

If you're evaluating AI readiness across your portfolio, you've likely encountered a troubling pattern. CEOs present ambitious AI roadmaps. IT teams tout pilot projects. Yet six months later, you're seeing minimal value creation and growing concerns about "shadow AI," or employees using consumer tools for work tasks, creating compliance and security exposure.

Recent data reveals the core problem: 40% of people now use AI in their personal lives, with 8% using it daily (Gallup). But organizational adoption lags dangerously behind. This gap represents both acute risk and significant opportunity for portfolio value creation.

The firms closing this gap in 2026 will gain measurable competitive advantage. Those that don't will find their portfolio companies increasingly behind market leaders.

The 80/10/10 Rule Most Boards Miss

In a recent NextAccess webinar featuring AI transformation leaders from Novo Nordisk, UserTesting, and ServiceNow, every executive independently arrived at the same insight. Successful AI transformation is: 

• 80% about people and change management

• 10% about practical tactics

• only 10% about technology

This contradicts how most portfolio companies approach AI. The typical pattern:

• Leadership hears about AI's potential

• IT is tasked with finding use cases

• Six months developing a solution

• Six months implementing it

• A year later: one AI application nobody uses, significant spend, minimal value

Meanwhile, competitors have pulled ahead.

"Too many companies think of AI as a technology problem to solve," explained Michael Domanic, Head of AI at UserTesting. "It's not. It's a creative problem. We need to rethink the way we're working in our business with these new capabilities."

For PE/VC firms, this has immediate implications for how you allocate transformation capital and evaluate management teams. Companies starting with major IT implementations before upskilling their workforce waste a year and significant money. Competitive dynamics are moving too fast for this approach.

What to Look for in Your Portfolio Companies

1. A Dedicated AI Transformation Leader (Not a Side Project)

The single most important indicator of AI readiness is whether a company has designated a full-time AI transformation leader with real authority and budget. This is not the CIO's side project. They have enough on their plate.

This person needs six specific capabilities:

• A change management mindset

• Fluency in AI tools (hands-on, not theoretical)

• Broad business knowledge across functions

• Experience with zero-to-one transformation

• Creativity to identify non-obvious applications

• Superior communication skills, including listening to employees

As Matt Laessig, former COO of data.world (recently acquired by ServiceNow), emphasized: "The technology is abstract. The use cases are abstract. You need someone who can think creatively about applications across the business."

Red flag: If your portfolio company's AI initiative is owned by IT without business transformation ownership, expect limited returns.

2. Focus on Business Outcomes, Not Technical Outputs

Dr. Alicia Abella, who leads AI transformation at Novo Nordisk's U.S. operations, offered crucial guidance for 2026 planning: "Start by identifying two to three business objectives where AI can make a measurable difference, and align those objectives to measurable metrics: percent reduction in cycle time, increase in revenue, percentage improvements."

The distinction matters for board oversight. ‘We implemented AI in customer service’ is an output. ‘We reduced average resolution time by 35% and increased customer satisfaction scores by 12 points’ is an outcome you can track against targets.

Green flag: Management teams that link AI initiatives to specific KPIs the board already tracks: cycle time, revenue per employee, customer acquisition costs, Net Promoter Score.

3. Quick Wins Over Long Implementations

Laessig described finding AI opportunities by identifying "areas of our business where we have high velocity and high ROI but can't keep up with demand."

At data.world, his team built a custom RFI response tool using their existing knowledge base. Instead of a laborious manual process taking days or weeks, they generated high-quality responses in hours—a 20x speed improvement directly correlated to sales success.

Common high-ROI quick wins across portfolio companies:

• RFP/RFI response generation (20x speed improvement)

• Meeting summarization and action tracking

• Content creation and first-draft writing

• Customer support inquiry routing

• Internal documentation search

• Data analysis and reporting

These require minimal data infrastructure work and can demonstrate value within quarters, not years.

Red flag: If a portfolio company's first AI project is a 12-month implementation with no planned quick wins, reset expectations now.

The Governance Paradox

In regulated industries—pharmaceuticals, financial services, healthcare—executives often cite governance concerns as reasons to delay AI adoption. This is backwards.

Dr. Abella addressed what she calls the biggest myth: "The myth is that AI governance actually slows progress. But if it's done right, it can actually enable AI to scale more quickly, safely, and confidently."

At Novo Nordisk, they built a risk assessment framework evaluating each use case across dimensions including:

• Internal vs. external-facing

• Use of personally identifiable information

• Human oversight requirements

• Technical complexity

• Deployment model

Low-risk projects move quickly through approval. Higher-risk projects get appropriate oversight. This creates velocity, not bureaucracy.

Laessig offered a useful analogy: "Governance is like brakes in a car. Brakes aren't there to slow you down. They let you drive fast safely."

For PE/VC firms with portfolio companies in regulated industries, good governance frameworks are enabling assets, not constraints. The companies that spend six months building perfect governance before anyone can experiment will be six months behind competitors who started with ‘good enough’ guardrails and learned by doing.

The Data Excuse

Many portfolio companies delay AI initiatives claiming ‘our data isn't ready.’ This is a red herring that signals management teams are not ready to act.

Laessig's advice cuts through this paralysis: "Start with your highest-value business problem—one with high ROI potential. Identify the AI applications that could solve it. Then work backwards to understand what data you need and get just that data ready. Then rinse and repeat."

Don't let portfolio companies boil the ocean. Focus on a specific business problem and the relevant data. There's too much value available from general-purpose AI tools (like the quick wins above) to wait for perfect data infrastructure.

Three Critical Myths to Challenge

Myth 1: ‘We can't trust AI because it hallucinates’

Yes, large language models can produce inaccurate information. But as Domanic noted: "If hallucination is holding you back from pursuing real AI transformation, you're thinking about AI completely the wrong way. Compare it to inaccuracies from human error. Is the hallucination rate greater than human error? Probably not."

The solution: appropriate verification based on use case risk, not avoiding AI entirely.

Myth 2: ‘AI will replace our employees’

Scott Kosch, CEO of NextAccess, offered a data-backed perspective: "If you look at the empirical data on who benefits most from using AI, the people who already are highly skilled only see a small bump up in their performance. It's actually people who have a lot of domain knowledge, but not necessarily some core fundamental skills, that see huge advancement."

Laessig's framing was also powerful: "Nobody here is going to lose their job to AI. But people will lose their jobs to humans who know how to use AI."

Myth 3: ‘We need a major implementation before seeing results’

The traditional consulting approach—spend six months scoping, six months building, six months implementing—wastes precious time. A better approach:

• upskill your workforce broadly

• enable experimentation within governance guardrails

• identify quick wins

• demonstrate value

• build momentum

Your Due Diligence Checklist

When evaluating AI readiness in portfolio companies, ask:

Immediate indicators:

• Is there a dedicated, full-time AI transformation leader?

• Have they defined 2-3 specific business objectives with measurable metrics?

• Do they have a basic risk assessment framework enabling safe experimentation?

Quarterly progress:

• Are they launching 1-2 quick wins per quarter with measurable ROI?

• Is the CEO communicating AI strategy and progress in every board meeting?

• Are they building workforce capability (training, ambassador programs) in parallel with technology investments?

Red flags:

• AI initiatives owned solely by IT without business transformation leadership

• First project is a 12-month implementation with no interim value demonstration

• ‘Waiting for data to be ready’ as reason for inaction

• No governance framework (or governance so rigid it blocks all experimentation)

The Window Is Closing

As Domanic emphasized "I think it is the responsibility of employers to lead their workforce through this transformation. We are living through one of the most transformational moments in history."

For PE and VC firms, 2026 is the year to stop planning and start executing across your portfolio companies. The transformation leaders featured in the NextAccess webinar didn't wait for perfect conditions. They started with what they had, focused initially on their people, built momentum with quick wins, and learned as they improved.

This article is based on insights from "Beyond the Hype: Your Practical 2026 AI Game Plan," a NextAccess webinar featuring AI transformation leaders from Novo Nordisk, UserTesting, and ServiceNow. For more information about this webinar and to schedule a complimentary 30-minute consultation to explore how NextAccess can help your organization, please message Scott Kosch or Valerie VanDerzee.

NextAccess Authors: Scott Kosch and Valerie VanDerzee

NextAccess is an advisory firm of experienced operators with deep experience running top-performing organizations and delivering exceptional results. We help executive teams and investors build stronger, more valuable companies through a powerful mix of operational expertise, strategic insight, and data-driven solutions.