Meta shares jumped 15% for the week, the best weekly gain since early 2024, as investors reassessed the company's AI roadmap following a flurry of product releases and a Reuters report that its custom chip program is advancing. The rally erased year-to-date losses, pushing the stock just above flat for 2024, though it still trails the Nasdaq's 13% rise.
New AI models target creators and enterprise workloads
Three months after introducing Muse Spark, its first proprietary foundation model, Meta announced two follow-ups this week. On Tuesday, the company released Muse Image, a model designed to generate images and attract creators and advertisers to new subscription offerings. Thursday brought Muse Spark 1.1, tuned for agentic and coding tasks. The releases signal Meta is moving quickly to close the gap with OpenAI, Anthropic, and Google, all of which have longer track records in commercial AI models.
The announcements underscore a strategic shift that executives tracking AI for Executives & Strategy will find familiar: the push to monetize AI through subscriptions, advertising tools, and cloud services. Meta's efforts also point to progress at its Superintelligence Labs, led by Alexandr Wang, as the company works to diversify revenue beyond its core ads business.
Custom chips and data center expansion
A Reuters report that Meta expects to begin manufacturing its first in-house AI chip, code-named Iris, in September added fuel to the stock's climb. The chip program, first disclosed in March as part of data center expansion plans, is central to Meta's goal of reaching 14 gigawatts of computing power next year. Bank of America analyst Justin Post said the move could reshape cost assumptions. "Meta may have engineered significant cost savings to get capacity cost per MW well below our and Street expectations," he wrote.
Wall Street warms to Meta's AI spending
When Meta reported first-quarter earnings in April, it raised 2026 capital expenditure guidance to as high as $145 billion. The stock dropped 7% that day as investors balked at spending that had yet to produce new business lines. Now, the tone has shifted. The chip progress and model launches give analysts a clearer picture of how the infrastructure will be used, including the possibility of competing in cloud computing against Amazon and Microsoft.
Nick Jomes, senior analyst at BNP Paribas Equity Research, said Meta could lift its 2026 capex guidance further when it reports second-quarter results, with his firm's estimate now in a range of $135 billion to $155 billion. He added: "While we expect near-to medium term elevated capex, we believe Meta is well positioned to generate ample revenue to support its spending, driven by monetization of its own AI initiatives, advertising share gains, incremental subscription revenue, an optionality of cloud offering, and fees for external use of its AI models."
Why this matters for executives
Meta's week shows that concrete product roadmaps and cost-saving hardware progress can quickly reverse investor skepticism about heavy AI spending. For strategy leaders, the takeaway is that the market is now pricing in not just the size of capex budgets but the evidence of a plan to turn them into revenue. Custom silicon could also lower the barrier for Meta to enter cloud services, a move that would reshape competitive dynamics for enterprise AI workloads.
Your membership also unlocks:
