3 Top Artificial Intelligence (AI) Stocks to Consider Buying in March
It’s been a challenging period for artificial intelligence (AI) stocks lately, with some industry leaders experiencing price pullbacks in the last month or so. However, this market volatility is also creating potential buying opportunities for astute, long-term investors.
Let’s examine three AI stocks that present such opportunities.
- Nvidia
Despite consistently reporting impressive revenue growth, Nvidia (NASDAQ: NVDA) has not escaped the recent market downturn. As of this writing, the stock price is down nearly 25% from its all-time high reached in January. Nevertheless, Nvidia remains exceptionally well-positioned to gain from the ongoing expansion of AI infrastructure, where spending is increasing this year.
Cloud computing companies are leading the way. Amazon (NASDAQ: AMZN) plans to invest approximately $100 billion in capital expenditures (capex) this year to build out its AI data centers. Microsoft is following with $80 billion, and Alphabet is planning $75 billion. Other companies are also aggressively pursuing AI opportunities. Meta Platforms is looking at up to $65 billion in capex this year, primarily for AI infrastructure, while a consortium spearheaded by OpenAI and SoftBank intends to spend $500 billion over the next few years on AI infrastructure through Project Stargate.
Nvidia is at the forefront as a leading designer of graphics processing units (GPUs). These are essential for both training AI models and for running inference. The company has captured a 90% share of the GPU market with its CUDA software. CUDA was the first platform created years ago, allowing developers to program these chips for purposes beyond their initial design of accelerating graphics rendering in video games.
Since then, Nvidia has developed an extensive collection of AI libraries and microservices, setting the company apart from competitors. While some companies have developed custom AI chips, they typically use them in conjunction with more flexible and readily available GPUs.
The stock’s recent downturn has positioned 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) ratio of under 0.5. PEGs below 1 generally suggest a stock is undervalued.
- Amazon
While recognized for its e-commerce operations, Amazon is fundamentally a technology company. The primary contributor to its profitability is its cloud computing unit, Amazon Web Services (AWS). The company pioneered the infrastructure-as-a-service model and remains the largest cloud computing company by market share.
AWS has also been the corporation’s fastest-growing segment, with revenue increasing by 19% to $28.8 billion last quarter. Its operating income surged 47% to $10.6 billion during the same period.
The segment’s strong operating leverage benefits from Amazon’s development of its custom AI chips called ASICs (application-specific integrated circuits). These are purpose-built for particular tasks and often outperform GPUs in these applications while consuming less power. However, they lack the versatility of GPUs.
Amazon reportedly handled the bulk of the work in designing these chips while licensing some technology from Marvell. This strategy provides the firm with a cost advantage.
Customers utilize AWS and its associated services to build their own AI models and applications. The company provides them with several leading foundational models to start with. Customers can utilize its SageMaker platform to customize and train their AI models before deploying them into production.
AWS has been capacity-constrained and will invest $100 billion in AI data centers this year to help meet the increasing demand. Amazon has a history of investing heavily to secure a strong market position, and it is leading the charge in AI.
The stock price is currently off by about 15% from its recent highs, and it trades at an appealing forward P/E of 32.
- Salesforce
Shifting to the software sector of AI, Salesforce (NYSE: CRM) is a solid option for investors. The customer relationship management (CRM) software company’s stock is down 20% from its recent highs. Salesforce aims to be a leader in agentic AI—that is, AI agents able to perform assigned tasks with minimal human oversight.
The company launched its new agentic AI offering, Agentforce, in October 2024. It has seen a strong initial acceptance, with Salesforce stating it has finalized 5,000 deals, including 3,000 that are already in the paying stage. Agentforce provides customers with several ready-made AI agents or allows customers to customize their own through no-code and low-code tools.
Earlier this month, Salesforce unveiled a new Agentforce marketplace, AgentExchange, with over 200 initial partners offering hundreds of ready-to-use solutions and templates to extend the potential applications of AgentForce. Several prominent software firms have already joined the marketplace, including Workday, Docusign, and Box.
Agentforce charges $2 per interaction, indicating substantial opportunity for the company as it gets more agents in operation. Ultimately, if Agentforce boosts productivity and cuts expenses for its customers, it should drive significant growth for Salesforce in the coming years.
The stock is attractively valued at 26 times 2025 analyst earnings forecasts and a 0.35 times PEG.
