2 Top AI Growth Stocks to Buy in September 2025: Your Portfolio’s Next Power Move

AI stocks surge as September buying window opens—smart money positions for Q4 rally.

First Pick: The Automation Architect

This company's neural networks process data 40% faster than legacy systems—cutting operational costs while competitors still run spreadsheets. Their proprietary algorithms already dominate supply chain logistics, with Fortune 500 clients scrambling for implementation slots.

Second Pick: The Cognitive Cloud Player

Enterprise adoption hits record highs as their AI-as-a-service platform bypasses traditional IT bottlenecks. Revenue growth accelerates while old-guard software vendors still debate cloud migration strategies—because nothing says innovation like charging monthly subscriptions for what used to be one-time purchases.

Both stocks trade at reasonable multiples despite crushing earnings estimates—because Wall Street analysts would rather downgrade for missing guidance by 1% than celebrate beating it by 20%.

Image source: Getty Images.

1. Broadcom

Broadcom has emerged as a key player in AI infrastructure, with a diversified portfolio of custom AI accelerators, advanced networking solutions, and enterprise software. The company serves several hyperscale customers and large AI model developers, providing them with the custom AI chips and networking technologies required to build large-scale AI clusters.

The numbers highlight the success of Broadcom's business model. In the third quarter of fiscal 2025 (ended Aug. 3), Broadcom reported revenue of $16 billion, up 22% on a year-over-year basis. AI semiconductor sales ROSE 63% year over year to $5.2 billion.

A major driver of this growth is the company's expanding custom accelerator or XPU business, as hyperscalers have been increasingly opting for custom AI silicon to improve performance and optimize costs. Demand from its existing three hyperscale customers continues to soar. Broadcom expects each of these three customers to deploy 1 million AI clusters by 2027, with a significant portion of these AI clusters involving custom accelerators.

Broadcom has also secured more than $10 billion in AI rack orders from a fourth hyperscaler customer, most likely OpenAI, and expects to commence shipping in 2026.

Networking has also become a significant catalyst for growth. As hyperscalers build massive AI clusters, they are increasingly facing challenges with speed, latency, and power efficiency. Broadcom's Ethernet switches and fabric routers are designed to address these bottlenecks, enabling efficient connection and network congestion management within data centers, as well as across data centers located on multiple sites.

By leveraging open Ethernet standards, Broadcom's Tomahawk switches and Jericho fabric routers provide higher flexibility at lower costs compared to proprietary networking solutions. Hence, Broadcom is well positioned to capture an even larger share of the AI networking market.

The company's enterprise software segment is also gaining momentum. VMware Cloud Foundation (VCF) 9.0, launched this year, enables enterprises to run AI and other workloads securely on private or hybrid cloud environments at lower costs. Software revenue grew 17% year over year to $6.8 billion in the third quarter.

Broadcom's stock is not inexpensive, trading at approximately 36 times forward earnings. However, the valuation seems justified considering the company's leadership in custom AI compute, networking, and enterprise software. With a solid backlog of $110 billion and an expanding customer base, Broadcom looks well positioned as a long-term AI growth story.

2. Meta Platforms

Meta Platforms is no longer just a social media company or a digital advertising giant. By combining an unmatched user base with massive AI investment, the company has rapidly transformed itself into a global AI powerhouse.

In the second quarter of fiscal 2025 (ended June 30), Meta's revenue climbed 22% year over year to $47.5 billion, while net income was a solid $18.3 billion. Operating margin expanded to 43%, and free cash FLOW reached $8.5 billion. The company returned nearly $11.1 billion through buybacks and dividends, while still ending the quarter with $47.1 billion in cash and $28.8 billion in debt.

Meta is aggressively leveraging advanced AI technologies to enhance content recommendations, user engagement, ad targeting, and ad monetization across its platforms, including Facebook, WhatsApp, Messenger, and Threads.

The company's AI-powered Andromeda ad retrieval models, generative ads recommendation system (GEM), and Lattice ad ranking models have helped boost ad conversion rates on Facebook and Instagram. Instagram Threads is also rapidly gaining traction, with more than 350 million monthly active users as of the end of the first quarter. Threads will soon become a solid monetization avenue for the company in the coming quarters.

Another significant competitive advantage is Meta's open-source Llama ecosystem. Llama models are now used widely by developers, who are often key decision-makers in enterprise technology adoption.

Meta has also planned to spend $66 billion to $72 billion in fiscal 2025 to expand its generative AI capacity by investing in servers, networking, and data centers. The company plans to increase capital investments in AI data center capacity even more in 2026. These investments in AI infrastructure capacity are expected to prove highly lucrative, as the company begins to increasingly utilize in-house AI capabilities at scale to enhance its Core businesses.

Despite the many pros, Meta faces significant risks. The company's digital advertising business is highly exposed to the overall state of the economy. The company also faces competitive pressures, while high AI-related capital expenditures can prove to be a drag on its profitability in the short term. But it is undeniable that the long-term prospects of scaling its AI infrastructure and Llama ecosystem remain attractive.

Meta trades at nearly 28.5 times forward earnings, which is not a cheap valuation. However, with strong financials, a growing moat built around its app ecosystem and AI infrastructure, and a commitment to returning significant value to shareholders, the stock appears to be a worthwhile buy now.