International Business Machines (IBM 1.89%) has found an artificial intelligence (AI) strategy that works. While the tech giant is an AI pioneer, monetizing the technology has been a challenge over the past decade. The company's most notable failure was an attempt to leverage AI to transform the healthcare industry. After spending billions on acquisitions to fuel the effort, IBM eventually abandoned the business a few years ago with little to show for it.
Today, IBM has settled into a role of enabling other enterprises to train, deploy, and manage AI models and agents. Through its watsonx AI software platform and its consulting arm, IBM offers a wide range of AI-related services that resonate with its client base.
When IBM reported its fourth-quarter results on Wednesday, the company noted a big acceleration in its generative AI business. IBM won nearly $2 billion worth of new AI business during the quarter, pushing the cumulative total to $5 billion.
NYSE: IBM
Key Data Points
Benefiting from disruption
The big news in the AI industry in recent weeks has been the revelation that Chinese start-up DeepSeek had managed to train a generative AI model that's capable of matching the performance of top-tier models from U.S. companies like OpenAI for a fraction of the cost. Not only did the model cost very little to train, but it's dramatically cheaper to run as well.
This development throws into question a lot of the assumptions that are driving mega-investments in AI infrastructure and booming demand for powerful AI accelerators. If powerful AI models can truly be trained cheaply and run on less powerful hardware, the AI industry is in for a massive and unpredictable shakeup.
IBM is not in the business of developing expensive, general-purpose AI models like some of its peers. Instead, the company provides a platform, expertise, implementation, and other services to help clients apply generative AI to real-world use cases. IBM's Granite family of AI models are small and cheap, designed for specific purposes. IBM's watsonx platform also supports AI models from other providers, giving customers a wide range of options.
For IBM, cheap and plentiful AI is a good thing, because fast models running on cheap hardware will likely expand the market for AI. Enterprises that were turned off by the cost of running the most capable AI models may give the technology a second look. If the ongoing cost of deploying a customer service AI bot were to drop tenfold, demand for all the services and software required to get it up and running could soar.
One important thing to know about IBM's AI strategy is that it's cloud agnostic. Clients can run watsonx on IBM's cloud, but they can also deploy the platform on AWS, Azure, or their own hardware. IBM's consulting arm, through strategic partnerships with cloud providers, can construct solutions for clients that are already committed to AWS or Azure. This openness to competing cloud platforms makes IBM a far more attractive choice for enterprises looking to deploy AI-powered applications and agents.
Buy this enterprise AI leader
IBM does a lot more than AI, and its growth in 2024 was driven by multiple lines of business. IBM's software segment grew revenue by 9% last year, adjusted for currency, with strong growth for transaction processing, Red Hat, and automation. Most of IBM's AI bookings come from consulting signings, although the company has surpassed $1 billion in bookings for AI software.
In 2025, AI will be one of many growth drivers as the company aims to accelerate its revenue growth. IBM expects to grow total revenue adjusted for currency by at least 5% this year, up from 3% in 2024. IBM is also predicting free cash flow of around $13.5 billion in 2025, up from $12.7 billion last year.
The era of cheap and powerful AI could very well be upon us, and IBM is poised to benefit. The company's AI business is accelerating, and as the cost to deploy and run capable AI applications plunges, demand for IBM's AI software and services could soar. While there's a lot of uncertainty in the AI industry right now, IBM looks likely to emerge as a big winner.
