Security, AI, and EVs: One Infrastructure Story

The four stories logged on this desk in the past two days - Tesla soliciting Cybercab fleet operators, a U.S. senator asking the NSA for VPN guidance, Accel reportedly leading a $1 billion round for Thinking Machines, and Zscaler beating Q4 estimates - look unrelated at first glance. Read together, they show a single pattern: the United States is increasingly relying on private companies, not public agencies, to construct and operate its most consequential digital and physical infrastructure. The EV charging network, the AI compute layer, and even the tools for secure communication are all being designed, funded, and run by corporations. For U.S. technology companies, the market, and consumers, that shift carries both opportunity and fragility.

Infrastructure Is No Longer Public

When a senator asks the National Security Agency for guidance on virtual private networks, the implicit admission is that the federal government no longer sets the standard for secure communication. As Ars Technica reported, the array of options - open source, commercial, single-hop, multi-hop, mixnet - is dizzying. That variety exists because private vendors, not state agencies, have built the market. The senator's request is a recognition that Washington is now a consumer of security infrastructure, not its primary architect.

The same dynamic appears in transportation. Tesla is not asking regulators to plan a robotaxi network; it is publishing a web form, as TechCrunch reported, soliciting people interested in purchasing Cybercab fleets. The company is effectively crowdsourcing the buildout of a new mode of public transit. No municipal authority has ordered these vehicles or set their routes. Instead, individual buyers will decide where fleets operate, how many vehicles to deploy, and what to charge riders. That is a fundamentally private approach to what has historically been a public utility.

Investors Are Betting on Private AI Scale

The reported $1 billion round for Thinking Machines, led by Accel, is another example. As TechCrunch noted, the startup's annual revenue run rate exceeds $100 million. That kind of valuation - $40 billion against roughly $100 million in revenue - reflects an expectation that this private company will become essential infrastructure. Thinking Machines is not selling a feature; it is selling the foundation on which other companies will build their AI products. The investors are not just betting on one firm; they are betting that the entire U.S. economy will rely on a handful of private AI providers.

Public institutions are not absent from AI, but they are followers. The NSA, which historically developed cryptographic standards and secure computing methods, is being asked to endorse or explain tools that private companies have already shipped. The federal government may regulate AI, but it does not build the models, the chips, or the data centers. Those are being constructed by companies like Thinking Machines, backed by venture capital from firms like Accel. The scale of that private investment dwarfs what any public agency could allocate.

Security Is Becoming a Subscription, Not a Standard

Zscaler's results, as SiliconANGLE reported, point to a similar trend in cybersecurity. The company beat Wall Street's revenue and earnings estimates for its fiscal fourth quarter and issued fiscal 2027 earnings guidance above analyst expectations. Shares still fell 2% after-hours, but the underlying numbers show that businesses are paying handsomely for cloud security delivered as a service. Zscaler is not a federal lab releasing guidelines; it is a vendor selling access to a security platform. Its customers include thousands of enterprises that have chosen to outsource part of their network defense.

The senator's VPN question and Zscaler's earnings are two sides of the same coin. In the 1990s, a U.S. senator might have asked the NSA to develop a recommended encryption standard for government use. Today, that senator asks for guidance on a bewildering array of commercial products. The NSA's historical role was to set the standard; now it is to help citizens navigate a marketplace. Meanwhile, Zscaler's business model depends on organizations treating security as a recurring operational expense rather than a one-time procurement. That is a profound shift from the era of government-defined perimeter defenses.

📣

728x90

MID_CONTENT_2

What This Means for U.S. Companies and Consumers

For U.S. technology companies, the pattern creates enormous opportunities. The demand for private AI infrastructure, EV fleet management, and cloud security is global, but the supply is concentrated in American firms. Tesla already dominates EV sales in the U.S. and is now trying to own the operating layer for robotaxis. Zscaler has carved out a profitable niche in cloud security. Thinking Machines, if the reported round closes, will have billions of dollars to hire talent and build data centers. None of these companies need a federal contract to succeed; they sell directly to consumers and enterprises.

For U.S. consumers, the consequences are mixed. On one hand, they get faster innovation. Tesla can deploy a Cybercab test fleet in a city without waiting for a five-year municipal pilot. Zscaler's software updates continuously, so threats are patched in days, not months. VPN options abound, from open-source tools to commercial multi-hop services. On the other hand, consumers now depend on the solvency and good behavior of private companies for services that were once public responsibilities. If Tesla's fleet operators fail to maintain their cars, a city's transportation options could collapse. If Zscaler suffers a breach, thousands of its customers are compromised simultaneously. If Thinking Machines goes bankrupt, every AI startup that built on its platform faces an existential crisis.

The senator's request for NSA guidance highlights a related risk: the federal government is so far behind the private sector that it cannot even recommend a VPN without consulting an agency whose original mission was signals intelligence, not consumer advocacy. That is not a criticism of the NSA; it is a reflection of how quickly technology outpaces governance. The NSA might eventually publish a guide, but by then the VPN market will have evolved. The mixnet that was obscure today might be obsolete next year.

watch: The Reliability Question

The thread running through these stories is not just privatization; it is the question of what happens when private infrastructure fails. Tesla's Cybercab form, as TechCrunch reported, asks for information from potential fleet buyers, but it does not promise insurance, maintenance standards, or liability coverage. Thinking Machines' $40 billion valuation, as reported by TechCrunch, rests on revenue of just over $100 million, an unusually high price-to-sales ratio even for a fast-growing software company. Zscaler's shares fell despite beating estimates, suggesting that investors are already asking whether the security market's growth can justify its premium multiples. The senator's request to the NSA implies that even the government is unsure which private security tool is trustworthy.

None of these facts point to an imminent crash; they point to a structural dependency. The U.S. and its consumers are wholeheartedly adopting a model in which the most critical systems - transportation, computation, communications security, and the AI layer that increasingly powers all three - are owned by a small number of public companies and venture-backed startups. Those companies are accountable to shareholders more than to citizens. They may be more efficient than public agencies, but they are not guaranteed to be more resilient. If a Cybercab fleet operator goes under, if Thinking Machines' model fails a safety test, if Zscaler's platform is breached, the government will not have a fallback plan. The senator's call to the NSA is a small acknowledgment of that gap, but the stories from this week suggest the trend is already past the point of reversal. The private sector has built the infrastructure; the public is buying it.

What to Watch

What matters next is not which company's stock moves but whether these private infrastructures prove reliable under stress. Watch whether Tesla releases details on maintenance and insurance for its Cybercab fleets, as the form on its website suggests that will be a deciding factor for buyers. Watch whether Thinking Machines' high valuation attracts scrutiny about its $100 million revenue run rate, since that gap could widen as more investors pile in. Watch whether the NSA, in response to the senator, endorses a broad category of VPNs or singles out commercial products, because that will signal how the government now views the private security market. And watch whether Zscaler's flat after-hours share price is a blip or the start of a broader recalibration of security valuations. All four watches are, at their core, tests of the same thesis: that America's future infrastructure is too important to be left to any single institution, public or private, but it will be built by the latter.

More on this beat: Companies on TechManNews.

📣

728x90

IN_ARTICLE_5

#AI infrastructure#EV fleets#Cybersecurity#Privatization#Tech policy#Venture capital

← Previous Post

When AI Ambition Meets Public Backlash, the Giant Stumbles

The Patch Is Dead; Long Live the Shield

Article

AI Money and Muscle Reshape Tech's Old Power Balance