LIVE FROM MWC25 LAS VEGAS: IBM’s Emily Fontaine (pictured, right) provided a roadmap for AI investments which included five key pillars as enterprises ramp up their deployments.

To set the stage for the pillars, Fontaine, global head of venture capital at IBM, provided an overview of just how explosive the AI sector is.

She explained enterprise adoption of AI has surged dramatically, increasing from 26 per cent a few months ago to 42 per cent, highlighting a critical trend toward tailored offerings across industries.

In the first half of 2025 alone, AI investment equalled $104 billion, surpassing all of 2024, reflecting a fast-moving and heavily funded ecosystem.

The AI market is projected to reach $5 trillion by 2033, representing a significant portion of global GDP.

“What’s exciting about that is it’s creating unprecedented opportunities for startups, for investors, for enterprises, for the entire ecosystem,” she said.

When it comes to investing in AI, the first pillar Fontaine examines is strategic alignment, which she defines as ensuring a long-term fit within organisational goals and technology trends.

“The next thing I look at is both technology and product,” Fontaine stated. “Do you have technology that’s both interesting, innovative and disruptive?”

The third pillar is market opportunity, which is assessing whether there’s an actual market for the company or partnership.

“We see a lot of interesting startups that pop up, but there’s no market yet for them,” Finley noted. “The next thing you’ve got to look at, and this is really important as you’re thinking about either investing or partnering, is the team.

“You’ve got to double down on the team. Do they have the technical expertise? When I say team, I don’t just mean the team of the developers and the founders or the demand-making team.”

The last pillar is financial discipline which includes confirming sound business models and financial structures. This approach helps differentiate startups that are genuinely driving value versus those merely experimenting with AI.

She recommends live demonstrations using real data to verify true AI capabilities.

“The question I’m asking people is, are you experimenting with AI or are you executing with AI? The next question, is have you done POCs?  That’s where you start to see the folks who are experimenting, the folks who are executing, the folks who are becoming a fabric of an institution, the fabric of an enterprise,” she explained.