Truth or Clickbait: Going Beyond the Sensational Headlines on Enterprise AI

Does it sometimes feel like the world of AI is constantly resetting the clock? One day you’re under pressure to deliver value fast with your AI project. The next, a prominent study claims most projects deliver zero value. Then, a leading business publication explains that, actually, value doesn’t even matter right now. All while the technology itself is advancing at a dizzying pace.

In this blog, I’ll take a look at recent attention-grabbing headlines and share my perspective on what actually matters and what’s just clickbait — and what it means for you.

Headline #1: Most Companies Shouldn’t Have an AI Strategy

The Wall Street Journal recently argued that most companies shouldn’t have an AI strategy — a striking reversal of the common wisdom that every business needs one. Their reasoning: many organizations simply aren’t ready. Their data isn’t trustworthy enough, and their workforce doesn’t yet have the skills to make AI successful. In that sense, a formal AI strategy risks being more distraction than driver.

At Qlik, we just completed our third annual AI study, and the findings highlight why readiness matters. Data quality, availability, and accessibility remain the top barriers to AI adoption, closely followed by a lack of internal expertise. Yet despite these gaps, 69% of large enterprises reported having a formalized AI strategy, up from 37% in 2024 — a sign that many large companies are prioritizing AI strategically, even while readiness gaps remain.

What does this mean for your company? Having an AI strategy isn’t wrong — in fact, it can help ensure your organization is thinking about how to adopt AI effectively. The key is to make sure it goes hand-in-hand with building the foundations: strengthening data, developing AI literacy, and aligning technology to outcomes. Think of it this way: companies once built separate strategies for mobile and e-commerce before these initiatives eventually became embedded into core business strategies. AI is following the same arc. By focusing on readiness now, your AI strategy will be positioned to succeed as AI becomes increasingly integrated into how business gets done.

Headline #2: 95% of Enterprise AI Fails

We’ve all seen the headline: “95% of Enterprise AI Fails.” The stat went viral, originating from an MIT study, and no doubt many of you have read it. But the truth is more nuanced. The study actually found that 95% of AI projects produce zero measurable value, and even that conclusion is complicated by methodological quirks, small sample sizes, and other limitations the report itself acknowledges. As Albert Einstein once said, “Not everything that can be counted counts, and not everything that counts can be counted”…

Our new study provides additional context on this. We found that only 19% of large enterprises currently have a well-defined AI framework — low, but consistent with the adoption curve of new technologies historically. At the same time, 54% of organizations now assess AI on a case-by-case ROI basis, up from 28% last year, showing clear progress. And with agentic AI advancing rapidly, technology is evolving faster than many companies’ current capabilities, adding another layer of complexity to adoption.

The takeaway for you: don’t panic about the “95%” headline. Focus on the foundations first — high-quality data, appropriate AI methods, and clear metrics. The technology will keep moving fast, but a solid foundation will help ensure your AI efforts create real, measurable value.

Headline #3: Stop Worrying About AI’s Return on Investment

But wait, does ROI even matter right now? Not according to a new Wall Street Journal column, which declares that companies shouldn’t worry about AI’s ROI. The reason: measuring the impact of AI on business productivity is extremely difficult — and when we do try, we’re often looking at the wrong metrics.

This perspective comes from what technology leaders shared at a recent WSJ Leadership Institute Technology Council Summit. They noted that small efficiency gains don’t capture AI’s true value, and most projects are still in proof-of-concept stages designed to explore possibilities rather than deliver immediate returns. They emphasized focusing on a few high-potential initiatives while ensuring the organization has the right talent, governance, and data infrastructure to scale them effectively.

I largely agree with this view: early AI initiatives are about experimentation and building the organization’s capabilities for scale. That said, ROI remains a critical consideration — you still need to prioritize projects that align with business objectives and will ultimately deliver measurable value. The key is balancing innovation with disciplined measurement so that AI investments are both transformative and strategically sound.

The Bottom Line

AI headlines are designed to grab attention — but, as everyone knows, the reality is rarely as simple as the clickbait suggests. Just because facts are in conflict with each other does not make them less true, it just demonstrates complexity exists and there is not a simple answer. What matters most for enterprises is staying focused on the fundamentals: building strong data foundations, maintaining an open technology ecosystem, developing AI-ready talent, and aligning AI initiatives to clear business outcomes.