What you want from TV advertising: Full-screen, non-skippable ads on premium platforms.

What you get: "Your ad is on TV. Trust us."

Modern, performance-driven CTV gets your TV ads where you want with transparent placement, precision audience targeting, and measurable performance just like other digital channels.

TV doesn't have to be a black box anymore.

Get Started

The artificial intelligence gold rush just collided with geopolitics, and the collision has a specific address: a $500 billion data center complex being built in Abu Dhabi called Stargate.

Last Sunday, Iran's Islamic Revolutionary Guard Corps issued an explicit threat to destroy it.

The threat came in the context of the ongoing US-Iran conflict, with Iranian military officials warning that OpenAI's flagship infrastructure project — a joint venture backed by the US government, SoftBank, and Oracle — would be a legitimate military target if the US continues its operations against Iran.

This isn't background noise. This is the AI story of the week, and it has real implications for every New York business owner who has integrated AI tools into their operations over the last 18 months.

What Actually Happened

OpenAI's Stargate project, announced earlier this year, is a $500 billion infrastructure buildout designed to give the US a decisive lead in global AI capacity. The Abu Dhabi facility is the international anchor of that project — a massive server complex intended to process AI workloads at a scale the continental US alone can't currently match.

Iran's IRGC identified it by name and called it a legitimate military target.

The same week, a separate but related story broke: OpenAI, Anthropic, and Google announced a joint effort to combat Chinese firms illegally copying their AI models. The US and China are in an AI arms race that isn't metaphorical anymore — it's being fought with chip export controls, intelligence operations, and now, apparently, military threats against physical infrastructure.

Meanwhile, Anthropic quietly reported it has crossed a $30 billion annual revenue run rate — tripling from $9 billion at the end of 2025. The AI industry is no longer a promise. It's an enormous, rapidly concentrating sector of the global economy.

And it runs on physical servers. In specific places. That can be targeted.

Why This Matters to NYC Businesses

New York City has become one of the most AI-dependent small business environments in the country. Walk into any agency, law firm, marketing shop, contractor's office, or retail operation in any of the five boroughs and there's a good chance they're running on ChatGPT, Copilot, Claude, or some combination of AI-powered tools for drafting, customer service, scheduling, or operations.

That's a dependency most business owners built quietly, almost accidentally, over the last 18 months. The tools were cheap, sometimes free, and the productivity gains were immediate.

What's less obvious is that this dependency is concentrated. OpenAI, Anthropic, and Google collectively power the vast majority of AI tools that businesses actually use. All three run on data center infrastructure — physical buildings, physical servers, physical power lines — that can be disrupted by storms, by cyberattacks, or, as of last week, by military action.

The geopolitical risk is real and it's no longer theoretical.

Which Industries in NYC Are Most Exposed

Some sectors have built harder dependencies than others.

Legal and professional services. Small law firms and accounting practices that use AI for contract review, document drafting, and research are running significant workflow risk if a major AI provider goes dark, even temporarily. Most haven't built a backup.

Marketing and creative agencies. The AI writing and image generation tools that have replaced significant headcount at NYC agencies are overwhelmingly concentrated in OpenAI's ecosystem. If ChatGPT goes offline for 72 hours, some of these shops genuinely can't deliver client work on deadline.

Healthcare and mental health practices. AI scheduling, transcription, and administrative tools are increasingly embedded in how NYC private practices operate. The dependency is newer but it's growing fast.

Retail and e-commerce. AI-powered inventory management, customer service bots, and ad targeting are standard at this point for any NYC retailer doing meaningful volume online.

None of these businesses have disruption plans for AI outages the way they have disruption plans for power outages or flooding.

What The Smart Money Is Doing

The major AI infrastructure players aren't ignoring the concentration risk. They're spending at a scale that would be impossible to absorb for any individual business.

Eclipse, a venture capital firm backed by Cerebras, raised $1.3 billion last week specifically for AI infrastructure and physical defense plays. Nvidia-backed Firmus Technologies raised $505 million for data center construction. Microsoft announced a $10 billion investment in Japan as part of its Asia AI buildout.

The companies building AI infrastructure are diversifying geographically, hardening their facilities, and building redundancy. The question is whether the businesses using AI tools have done any of that planning at their scale.

Most haven't.

What NYC Business Owners Should Do Right Now

This isn't a reason to stop using AI tools. The productivity gains are real and abandoning them to competitors who keep using them is a worse outcome than occasional service disruption. But it is a reason to operate with some basic risk hygiene.

Audit your AI dependencies. Make a list of every AI-powered tool your business uses for critical functions. Note which underlying provider they run on. Most tools — even branded ones — run on either OpenAI's API or Google's. A failure at either of those companies cascades across hundreds of third-party applications.

Identify your fallback. For each critical AI function, what's the human or manual alternative if the tool goes dark for 24 to 72 hours? If the honest answer is "we don't have one," that's the gap to close.

Diversify where you can. If you're running entirely on OpenAI-powered tools, adding one Google or Anthropic-backed alternative to your stack reduces your exposure. This applies especially to tools you use daily for client-facing work.

Negotiate SLAs. If you're paying for enterprise AI tools, your contract should specify uptime guarantees and what happens when the provider fails to meet them. Most small businesses sign these agreements without reading them. This is worth knowing.

Don't build advertising strategy on AI-dependent channels alone. The most resilient marketing strategies use channels that don't require a live API connection to function. Connected TV advertising — platforms like Roku that let you reach specific audiences with pre-produced spots — operates independently of the generative AI infrastructure layer. If your marketing is entirely dependent on AI-generated content flowing through AI-targeted platforms, a disruption hits twice.

The Bigger Picture

The $122 billion OpenAI raised this year, the $30 billion Anthropic is now generating annually, the trillion-dollar buildout of physical AI infrastructure — these numbers signal that artificial intelligence has crossed from experimental technology to critical infrastructure. The same transition happened with the internet in the late 1990s and with cloud computing in the 2000s.

Every time a technology becomes critical infrastructure, it becomes a target. For cybercriminals, for competitors, and now, apparently, for state actors with military capabilities.

Iran threatening a data center in Abu Dhabi sounds distant. But the tools that data center powers are already inside your business. That's worth a few hours of planning before something forces the issue.

The Metro Intel covers AI, business, and real estate news for New York City's residents and business owners. If this was useful, forward it to someone who needs it.