Tech Corner: Is All This AI Spending Paying Off?

By Daniel Morgan, Synovus Trust Senior Portfolio Manager
Synovus Trust Company, N.A.

Meta’s (META) highly scalable (3.6 billion-plus daily average users) platform is evolving to support advertising, commerce and messaging, built on an AI-enabled infrastructure. Significant CapEx investments in AI are driving improved engagement and monetization for users/advertisers/creators. Meta AI assistant had 700 million users as of FY24, making it one of the most-used AI apps. Long-term opportunities include incremental monetization from Reels, WhatsApp/Messenger, Threads, ecommerce and Reality Labs/Metaverse. Key growth drivers include: 1. Positive impact from Advantage + AI Tools on Ad Growth and AI’s impact on engagement time for Reels versus TikTok; 2. Advertiser sentiment by vertical, with Ad Revenues in 3Q25 projected at $48.616 billion +21.9%; 3. Reality Labs spend and CapEx/OpEx outlook; 4. Potential for a $7 billion-plus revenue contribution for WhatsApp advertising; 5. For Threads, point to a $500 million-plus  revenue opportunity; and 6. Finally, valuations appear compelling, with Meta shares trading at 25x FY2025 EPS-GAAP estimate of $28.192  per share and 24x FY2026 EPS-GAAP estimate of $30. This compares to a five-year average P/E of 23x. So the shares appear fairly priced at this point compared to historical averages.

Meta has a series of AI-based ad tools/products that customers can use to maximize their digital advertising ROI. As advertisers each innovate to improve the ease of setup and management of ad campaigns, believe Meta is well positioned given the Advantage+ ad product, broad portfolio of consumer data and wide variety of surfaces available to advertisers. META has emphasized AI-integration into Advantage+ across several major AI-driven enhancements aimed at streamlining campaign management and improving performance. The company reported that AI-assisted identification of high-quality leads for advertisers reduced cost per qualified lead by 10%.

Similarly, by 1Q25, META saw strong adoption of its AI creative tools, with more than 4 million advertisers using Meta's generative AI ad creative tools. AI-based Ad tools have helped META accelerate the quarterly Ad revenue growth rate from the mid-teens to 20% plus.

For Microsoft (MSFT), investor sentiment has improved, pushing MSFT shares near all-time highs. Investors may still be underestimating the potential for Microsoft’s AI business to drive durable consumption growth for Azure and scale fast in the Agentic AI era given how underwhelming the value proposition and use cases for the first iteration of Copilot proved. Previously, the consensus view proved too bullish when Copilot first came to market on the adoption and monetization impact from Copilot. Microsoft’s success in Cloud is visible through the recent Azure revenue re-acceleration, the fast growth and size of its Azure business. The Azure business has a visible path to accelerating bookings at already an upper-echelon software industry growth rate (i.e., 30%+). Azure growth heading into the 1Q26 print is projected at 37.33%, with 18.71% PPTs coming directly from AI. The Intelligent Cloud segment is expected to also drive growth in 1Q26, with estimated revenues of $30.17 billion +25.2% Year-over-Year (YoY). Valuations appear compelling, with Microsoft shares trading at 33x FY2026 EPS-GAAP estimate of $15.61 per share and 28x FY2026 EPS-GAAP estimate of $18.272. This compares to a five-year average P/E of 33x.  So the shares appear fairly priced at this point compared to historical averages.

Currently, platforms generate almost twice as much revenue for the industry as ChatGPT and Copilot. For Microsoft, platforms will remain 3x over-the-top subscriptions as Azure will remain dominant hosting hub for both OpenAI and other models, including open source. Macro Environment appears mixed more inconsistent than usual, with geopolitical and trade restriction headwinds raising concerns about IT budgets. On the positive side, cloud infrastructure and GenAI investments remain strategic areas for Enterprises, but a sustained negative macro backdrop might lead Enterprises to reduce spending on growth initiatives, shifting their focus to “keeping the lights on.”

Concerns about reports of Microsoft canceling some datacenter leases could be overdone, given its massive investments in infrastructure, and could partially be attributed to the company’s success in ramping its own infrastructure. Azure growth heading into the 1Q26 print is projected at 37.3%, with 18.0% PPTs coming directly from AI.

Recently, OpenAI gave its long-time backer Microsoft Corp. a 27% ownership stake as part of a restructuring plan that took nearly a year to negotiate, removing major uncertainty for both companies and clearing the path for the ChatGPT maker to become a for-profit business. Under the revised pact, Microsoft will get a stake in OpenAI worth about $135 billion. In addition, Microsoft will have access to the AI startup’s technology until 2032, including models that achieved the benchmark of artificial general intelligence (AGI). OpenAI had spent much of this year working to restructure as a more traditional for-profit company. Microsoft, which backed OpenAI with $13.75 billion, was the biggest holdout among the ChatGPT maker’s investors terms of the contractual agreement. This is extremely important as it clarifies the relationship between MSFT and OpenAI and creates a path to profitability for Open AI.  As of the first half of 2025, OpenAI generated $4.3 billion in revenue but also reported a significant net loss of $13.5 billion.  OpenAI's losses will continue to burden Microsoft's EPS until it achieves profitability, or it reaches the $13 billion Market Cap (overall value of the investment), which expects to occur at the end of FY26. As such, estimates point to Microsoft's EPS growth accelerating in FY27, as the OpenAI other income loss rolls off once Microsoft has amortized the entirety of its initial investment. Further, the future strength of the OpenAI/MSFT relationship had come under some question as Open AI struck a deal with MSFT’s (Azure) competitor, Oracle (OCI), in the datacenter space. This agreement helps solidify the MSFT/OpenAI relationship.

Alphabet (GOOGL) fundamentals are still solid and the company will remain both a driver of and primary beneficiary of an increasingly digital economy and advances in Generative AI. Google remains focused on innovation and expects there is a healthy runway across Search and YouTube ads as AI drives higher ROI and TV dollars shift online. Non-ad businesses – including Cloud and YouTube subscription services – have significant headroom. And companies within Other Bets – including Waymo and Verily – provide option value.  Coming into this 3Q25 print, Google’s strong innovation quietly continues while macro issues, tariffs, U.S. Department of Justice and AI search competition capture the headlines.  Key growth drivers – advertising revenues are projected to grow to $72.337 billion (+9.84%) and Cloud (GCP) is expected to post revenues of $14.712 billion (+29.6%).

Alphabet has been working on its own custom AI chips for years. The company's Tensor Processing Unit, or TPU, was first announced in 2016. TPUs are optimized for tensor operations and deep learning, offering high performance and efficiency for those tasks, while GPUs (produced by Nvidia and Advanced Micro Devices) are more versatile and suitable for a broader range of computations. GOOGL is in the process of creating the next gen v6 TPU at 3nm, which is expected to reach production later this year. GOOGL may someday rely on the Tensor Processing Unit to power Google Cloud.  Alphabet broke new ground in April, with the release of the new CPU Axion. Google Axion is a family of custom Arm64-based processors designed for general-purpose computing, offering improved performance and energy efficiency compared to existing Arm-based instances and is underpinned by the Titanium system. The CPU Axion was built using the Arm Neoverse V2 CPU; Axion processors deliver giant leaps in performance for general-purpose workloads like web and app servers, containerized microservices, open-source databases, in-memory caches, data analytics engines, media processing, CPU-based AI training and inferencing, and more. 

The investment community has been worried that products like Open AI/ChatGPT could be having an impact on Google’s Search query growth. AI Search competition remains front and center given the rapidly growing scale of Open AI/ChatGPT usage and Google’s more measured pace in disrupting its own search ecosystem. New data disclosed by Google shows “over 5 trillion” queries annualized in January 2025. GOOGL management has stated repeatedly on its conference calls that “AI appears to be expanding the pie.” This fact has been heavily litigated among the investment community, which bears pointing to ChatGPT’s ascendancy to 800 Mn users every week (WAUs) as a potential risk. Search revenues represent 55% of total company revenues. For the upcoming 3Q25, Search is estimated at $54.8 billion (+11.2% YoY). Analysts will be focused on whether Search revenue growth is decelerating due to competition in queries from Open AI/ChatGPT. The jury is still out on how both Open AI/Chat GPT user growth will impact Google’s query growth.

For a period of time, the AI story was driven by this circular CapEx story at the top of the market, given that one firm’s CapEx is another firm’s revenue. Will the monster CapEx budgets announced by the hyperscalers yield the ROI that investors will demand to drive tech stocks higher? Of the top four hyperscalers – Amazon, Microsoft, Meta and Alphabet – they are collectively estimated to spend north of $342 billion (+42%) in CapEx in 2025 to build out their AI presence. If we add in Oracle’s recent CapEx guide increase to $30 billion, the combined figure is more than $370 billion. This compares to a combined CapEx estimate spend rate of $240 billion in 2024. (META 2025E CapEx guide to $66-72 billion; GOOGL 2025E CapEx guide to $85 billion; AMZN 2025E CapEx guide $100 billion; and MSFT 2025E CapEx guide  $87 billion).

Time will bear out whether all this CapEx spending will actually lead to enhanced growth rates in revenues/earnings. Currently, there is only a small amount of evidence pointing to actual increases in growth in various segments as a direct result of AI (i.e., MSFT’s Azure growth or META’s ad revenue acceleration). Productivity benefits from AI appear to be something that would take a bit longer than the consensus expected but have remained positive on the potential for AI to increase margins and labor productivity for companies that can take advantage of it. Now, with the cost of inference falling, expect the more exciting part of this new compute platform to be developed into solutions that can be leveraged by a wider swath of companies driving higher profitability and margins.