It's been a tumultuous time for artificial intelligence (AI) stocks recently, with the stock prices of some of the leaders in the field pulling back over the last month or so. However, the recent market volatility has also created potential buying opportunities for long-term investors who know which fallen stock to pick.

Let's look at three AI stocks that have that potential and are worth a closer look this month.

1. Nvidia

Despite recently reporting another fabulous quarter of huge revenue growth, Nvidia (NVDA 1.92%) has not been spared from the recent market sell-off. As of this writing, the stock is now down nearly 25% from its all-time high hit in January. Nonetheless, it remains the company best positioned to benefit from the current race to build out AI infrastructure, where spending is on the rise this year.

Cloud computing companies are leading the way, with Amazon (AMZN -0.72%) targeting around $100 billion in capital expenditures (capex) this year for building out its AI data centers, followed by $80 billion from Microsoft, and $75 billion from Alphabet. These aren't the only companies chasing AI, with Meta Platforms looking at up to $65 billion in capex this year mostly related to AI infrastructure, while a consortium led by OpenAI and SoftBank is set to spend $500 billion over the next few years on AI infrastructure through Project Stargate.

Nvidia is the leading designer of graphics processing units (GPUs), which are used to train AI models and run inference. It has been able to take a 90% share in the GPU market with its CUDA software, the first platform created years ago to let developers program these chips for tasks outside their original purpose of speeding up graphics rendering in video games.

In the years since, it has built out an extensive collection of AI libraries and microservices that set it apart. And while some companies have turned to custom AI chips, they generally use them in conjunction with the more flexible and readily available GPUs.

The recent sell-off in the stock has left it at a very attractive forward price-to-earnings ratio (P/E) of 25.5 times 2025 analysts' estimates and a price/earnings-to-growth (PEG) of under 0.5 -- with PEGs below 1 usually indicating a stock is undervalued.

Image source: Getty Images

2. Amazon

While best known for its e-commerce operations, Amazon is a tech company through and through. The biggest contributor to its profits is its cloud computing unit Amazon Web Services (AWS). The company created the infrastructure-as-a-service model and remains the largest cloud computing company by market share.

AWS has also been the company's fastest-growing segment, with revenue climbing 19% to $28.8 billion last quarter, while its operating income surged 47% to $10.6 billion.

The segment's solid operating leverage is helped by Amazon having developed its own custom AI chips, called ASICs (application-specific integrated circuits). They are designed for very specific purposes and tend to outperform GPUs in these tasks while using less power. However, they lack the flexibility of GPUs.

Amazon reportedly did the heavy lifting in designing these chips, while licensing some technology from Marvell. This helps give the company a cost advantage.

Customers are using AWS and its services to help build their own AI models and applications. The company provides them with several leading foundational models from which to start. Customers can use its SageMaker platform to customize and train their AI models and then move them into production.

AWS has been capacity-constrained and will pour $100 billion into AI data centers this year to help meet demand. Amazon has a history of spending big to win big, and it is leading the charge in this area.

The stock is off about 15% from its recent highs and now trades at an attractive forward P/E of 32 times.

3. Salesforce

Turning to the software side of AI, Salesforce (CRM -1.10%) is a solid option for investors with the stock down 20% from its recent highs. The customer relationship management (CRM) software company is looking to become the leader in agentic AI -- that is, AI agents that can complete tasks assigned to them with little human supervision needed.

The company's new agentic AI offering, Agentforce, was launched in October 2024 and has seen a strong initial reception, with Salesforce saying it has 5,000 deals in place, including 3,000 that are already paying. Agentforce offers customers several out-of-the-box AI agents from which to start, or customers can create their own through no-code and low-code tools built into the platform.

Earlier this month, Salesforce launched a new Agentforce marketplace called AgentExchange where it has signed up more than 200 initial partners to offer hundreds of ready-to-use solutions and templates to expand potential AgentForce uses. A number of leading software companies have already joined the marketplace, including Workday, Docusign, and Box.

Agentforce costs $2 per interaction, so this is an enormous opportunity for the company the more agents it can get into the wild. Ultimately, if it can increase productivity and save customers' money, it should be a big growth driver for Salesforce in the years ahead.

The stock is attractively valued at 26 times 2025 analyst earnings estimates and a 0.35 times PEG.