IBM Is Winning Enterprise AI in 2025 – Here’s How

IBM may not be the first name that comes to mind when talking about artificial intelligence (AI), but it was one of the first companies to bring AI to the grand stage.

Its Deep Blue system introduced early AI technology to the world in 1997. Running on a bespoke chess-playing supercomputer, it became the first machine to defeat a world chess champion, Garry Kasparov.

Even in the modern AI era, IBM hasn’t lagged behind. In October 2020, the company famously announced a strategic shift, splitting into two entities – one focused on hybrid cloud computing and AI and the other on managed infrastructure services.

However, unlike companies such as OpenAI, Google, and Microsoft, IBM focuses its AI initiatives around practical business applications rather than positioning AI itself as the primary selling point.

In this article, we explore how this distinct approach sets IBM apart in the enterprise AI market.

Table of Contents Table of Contents

1. Key Takeaways
2. IBM Is Doing AI Differently
3. What Do the Numbers Say?
4. The Bottom Line
5. FAQs
6. References

Show Full Guide

Table of Contents

1. Key Takeaways
2. IBM Is Doing AI Differently
3. What Do the Numbers Say?
Show Full Guide
4. The Bottom Line
5. FAQs
6. References

IBM Is Doing AI Differently

IBM’s AI business has many layers, but two key strategies set it apart from the competition. First, IBM focuses on the business case for AI rather than trying to get customers excited about the AI product itself.

Second, instead of heavily investing in AI models like OpenAI and Google, IBM is taking a more targeted approach, building smaller AI models designed specifically for enterprise needs.

Focus on the Business Case

IBM aligns AI with specific business goals, like improving sales targeting, manufacturing efficiency, and overall operations.

AI has to make a real financial impact. Many AI use cases, like writing employee reviews or drafting emails, do not actually move the needle for a large enterprise.

Think about it this way – a company like JP Morgan Chase will not see major gains just because AI helps them write emails, press releases, or marketing copy. These are just minor use cases that will hardly make a big difference.

IBM gets this and has shaped Watsonx as a business intelligence and analytics tool that helps companies find where AI can actually deliver real value.

IBM has been listening to enterprises that already had data-driven initiatives in place. AI, for them, is a natural extension of business analytics – not a revolution, but an evolution that delivers tangible results.

Small Steps for Big Results

Training large AI models is expensive (unless you can work miracles with $6 million and old Nvidia GPUs). But massive investments are not always necessary to get big returns, at least for now.

IBM has built its own AI models specifically for enterprise use. The latest version, Granite 3.0, goes up to 8 billion parameters.

That might sound big, but it is small compared to the most powerful generative AI models, which have hundreds of billions of parameters. More parameters generally mean a more capable model, but IBM is taking a different approach.

Instead of chasing size, its models are fine-tuned by enterprise customers using their own proprietary data. A smaller, well-tuned model can outperform a much larger one in specific areas, making it a smart fit for business applications.

Going even further, the Granite 3.0 family includes smaller models built for low-latency applications that can run directly on a CPU without the need for expensive AI accelerators.

For enterprises looking to run AI on their own hardware, these models offer a more affordable option, eliminating the need for costly data center GPUs.

Granite 3.0 is our latest update for the IBM foundation models. The 8B and 2B models outperform strong competitors with similar sizes. The 1B and 3B MoE use only 400M and 800M active parameters to target the on-device use cases. Our technical report provides all the details you… pic.twitter.com/N9hsn1IXzs

— Yikang Shen (@Yikang_Shen) October 21, 2024

What Do the Numbers Say?

Let’s look at the numbers. In January 2025, IBM reported fourth-quarter earnings that beat Wall Street expectations for both revenue and profit. Obviously, not all this growth came from AI. It is still a small part of IBM’s overall business, but it is growing rapidly each quarter.

IBM’s software segment grew 10% year over year to $7.9 billion, helped by demand for AI and strong performance from its Red Hat Linux business.

Since its launch, IBM’s generative AI book of business has grown to over $5 billion, adding nearly $2 billion in just one quarter.

The Bottom Line

IBM is not chasing AI hype. It is taking a practical, business-first approach that aligns AI with real enterprise needs. Instead of pouring billions into massive models, it is building smaller, fine-tuned AI systems that deliver real value at a lower cost.

Its AI business is still growing, but the numbers show that enterprises are buying into its vision. While others focus on making AI bigger, IBM is making it work where it matters most – driving business impact.

FAQs