Tesla’s first-quarter results cleared Wall Street’s bar, but the stock still dropped 3.6 percent on April 23, 2026, after Elon Musk told investors the company plans to spend roughly $25 billion on artificial intelligence infrastructure and robotics this year. That figure, approximately triple what Tesla spent in 2025, turned what should have been a victory lap into a pointed debate about how much cash the company is willing to burn chasing a future beyond cars.
The earnings beat that nobody celebrated
Tesla reported revenue and adjusted earnings per share above consensus estimates for the first quarter, buoyed by steady deliveries of the Model Y and improved manufacturing costs. Analysts had lowered expectations earlier in the year after a string of global price cuts, so the bar was not especially high, but the company cleared it convincingly enough that the numbers themselves drew little criticism, according to earnings coverage from Yahoo Finance.
The mood shifted the moment Musk laid out the capex plan. On the earnings call, he described two main spending buckets: expanding data centers and custom compute hardware for AI model training, and scaling production of Optimus, Tesla’s humanoid robot. Musk argued that Optimus could eventually generate more revenue than vehicles, a claim he has made before but one that now carries a $25 billion price tag. Investor’s Business Daily noted that the capex announcement dominated the Q&A session, pushing questions about vehicle margins and delivery guidance to the margins.
Shares closed down 3.6 percent, a notable underperformance against large-cap tech and auto peers on a day when broader markets were already choppy amid shifting interest-rate expectations, as tracked in Motley Fool’s live market coverage.
Why investors flinched
The spending plan matters because it rewrites Tesla’s financial identity. For years, the company pitched itself as a capital-efficient manufacturer that could scale EV production without the bloated factory footprints of legacy automakers like Ford and General Motors. Tripling capex to build AI training clusters and robot assembly lines breaks from that story. Free cash flow, already under pressure from price cuts, will face further strain before any meaningful revenue from Optimus or AI software subscriptions materializes.
The strategy also sharpens a question that has divided Tesla bulls and bears for years: Is this a car company with tech upside, or a tech company that happens to sell cars? By pouring capital into custom AI chips, neural-network training infrastructure, and vision-based robotics, Musk is betting heavily on the second interpretation. If Full Self-Driving adoption accelerates and Optimus reaches commercial production, Tesla could unlock high-margin software and robotics revenue that looks more like enterprise tech than automotive manufacturing.
But the execution risk is steep. Tesla is now asking investors to believe it can simultaneously build world-class AI infrastructure, compete with specialized chip designers like Nvidia, industrialize a humanoid robot, and defend its position in an increasingly crowded EV market. A stumble in any one of those lanes could leave the company with billions in sunk costs and growth that falls short of the premium baked into its stock price.
Competitive dynamics add another layer of pressure. Traditional automakers are investing in driver-assistance technology, but none are matching Tesla’s AI spending levels. Meanwhile, hyperscalers like Amazon and Microsoft, along with Nvidia itself, are racing to build data-center capacity that overlaps with what Tesla wants to own in-house. By vertically integrating more of the AI stack, Tesla reduces its dependence on outside suppliers but also enters a capital-intensive arms race against companies whose core business is infrastructure.
Milestones that will settle the argument
Promises are easy at $25 billion; proof is harder. Over the next several quarters, investors will be watching a short list of concrete markers.
On the software side, the key metric is Full Self-Driving take rates. Higher adoption and measurable improvements in safety data would validate the argument that AI investment is generating recurring revenue per vehicle. Any new regulatory approvals for supervised or unsupervised autonomous driving on public roads would further bolster the case.
For Optimus, the threshold is moving beyond controlled demos into real pilot deployments. That could mean robots handling repetitive tasks inside Tesla’s own factories or limited trials with outside industrial customers. Evidence that Optimus can operate reliably, safely, and at a cost that justifies replacing human labor would do more to justify the capex plan than any earnings call slide deck.
Until those milestones arrive, Tesla’s stock is likely to trade less on quarterly delivery numbers and more on whether Musk’s most expensive bet yet is starting to pay off.
