Anthony Di Pizio, The Motley Fool
6 min read
After posting blistering gains for the past three years, many of the stocks leading the artificial intelligence (AI) revolution are pulling back. Investors are worried about a range of potential risks, from rising funding costs for data centers to weaker future demand than anticipated, which could derail the infrastructure spending boom.
The iShares Semiconductor ETF, which invests in the leading suppliers of AI chips and components, is currently trading in bear territory after declining by more than 20% from its June peak. The downturn is starting to weigh on major indexes like the Nasdaq-100 and the S&P 500, which could hurt investors’ sentiment toward the market overall.
Missed AI’s “Act 1”? Act 2 Could Be 14x Bigger. Most investors think they missed the AI boat because they didn’t buy Nvidia in 2005. But according to our analysts, we’re only at the end of “Act 1″—the R&D phase. “Act 2” is the global rollout. Continue »
I have my eye on two AI stocks from outside the semiconductor sector that could be great buys if the sell-off in the broader market intensifies. Here they are.
1. Datadog: Providing critical tools for AI deployment
Tens of thousands of businesses use Datadog‘s (NASDAQ: DDOG) cloud observability platform to monitor their digital infrastructure, because it can immediately alert them to technical glitches that might be affecting their sales channels or mission-critical software applications. I’ve advised caution on this stock during the past few months solely because of its high valuation (which I’ll discuss further in a moment), but a correction could be a great long-term buying opportunity.
During the past couple of years, Datadog has released a suite of observability tools specifically for the AI industry. LLM Observability, for example, helps developers identify technical issues, track costs, and analyze the quality of outputs when building large language models (LLMs). These models sit at the foundation of customer-facing chatbots and agents, so errors or inaccuracies could hurt a business both financially and reputationally.
In April, Datadog launched another product called GPU Monitoring to help businesses track usage costs and uncover technical issues when deploying graphics processing units (GPUs). These are the primary data center chips used when developing and running AI software, so they must be performing at optimal capacity to maximize processing speeds and minimize costs.
Datadog said all 10 of the AI industry’s leading companies were customers as of June 30, and although it hasn’t named them, we can reasonably assume the likes of Anthropic and OpenAI are among them. Datadog also said Model Context Protocol (MCP) server calls quadrupled during the second quarter of 2026, compared to the first quarter just three months earlier. In simpler terms, that means the amount of AI-related activity occurring in Datadog’s ecosystem is surging.
