The artificial intelligence (AI) industry is evolving rapidly. AI is getting smarter, although some cracks in the growth story have emerged. OpenAI‘s highly anticipated GPT-5 model has fallen well short of inflated expectations, which likely throws some cold water on the idea that AI “superintelligence” is right around the corner.

While uncertainty is running high, International Business Machines (IBM -0.85%) and Intel (INTC -0.59%) are two AI stocks worth buying if you have $1,000 available to invest. IBM is focusing on applying AI to real-world problems for its clients, and Intel could find itself manufacturing future AI chips for third-party customers if it can get its act together.

Here’s why IBM and Intel both look like solid AI bets.

Image source: Getty Images.

IBM: Using AI in the real world

Right now, companies like Nvidia that supply the powerful GPUs that make the AI boom possible are raking in most of the cash. In the long run, though, companies that take AI technology and provide high-value services will likely be the big winners. IBM is already doing exactly that.

An astonishing 95% of AI pilots at companies fail to produce results, according to an MIT report. With many businesses struggling to implement AI in a way that actually increases revenue or reduces costs, IBM already offers a solution. The company’s generative AI business has drummed up $7.5 billion worth of business so far, with most of that total coming from its consulting arm. AI implementation and integration services are proving to be critical pieces of the puzzle for enterprises.

As companies shift from frantically deploying AI to really thinking about returns on investment and real-world results, IBM is in a great position to grow its generative AI business. With the global economic environment growing more uncertain, AI projects that promise reduced costs and greater efficiency should be in high demand.

That’s exactly what IBM supplies. By pairing consulting with software, IBM has found a lucrative AI niche that can drive growth for years to come.

Intel: The foundry opportunity

There’s no question that Intel has largely failed to take advantage of the AI boom so far. The company’s AI accelerator efforts have been a disaster, with its Gaudi AI chips selling poorly and its next-generation Falcon Shores cancelled. New CEO Lip-Bu Tan has noted that it may be too late for Intel to succeed in the AI training market, which is dominated by Nvidia.

While Intel may not be a major player in the AI accelerator market, the future is likely to be less concentrated. Nvidia still sells the lion’s share of powerful AI chips, but many tech giants are already designing custom AI chips of their own, particularly for AI inference workloads. Reducing the costs associated with running powerful AI models is critical for companies like Alphabet, which is integrating AI into its core search business.

This explosion of custom AI chips could be a boon for Intel if the company can get its act together in the semiconductor foundry business. While the Intel 18A process is ready for volume manufacturing, the company has failed to win a major external customer. Intel 14A, which is expected to be ready in 2027, may be the company’s last shot at making the foundry business work.

If Intel can sell AI chip designers on its Intel 14A process, the company could turn its money-losing foundry bet into a highly profitable endeavor. The U.S. government’s stake in Intel could provide some incentive for companies to choose Intel for manufacturing, given the Trump administration’s push to boost U.S. semiconductor manufacturing. Even a single large customer win for Intel’s foundry could send the stock soaring as it finally taps into the booming AI chip market.