Datacenters are emerging as a new target in warfare.

Hello and welcome to TechScape. I’m your host, Blake Montgomery. If you find this newsletter engaging, please share it with someone you believe would enjoy it as well.

The US-Israel War on Iran: Datacenters as a New Frontier in Warfare

Iran has recently targeted datacenters in the Persian Gulf in an effort to dismantle symbols representing the technological partnership between the Gulf states and the United States. These datacenters, among the most costly structures globally, would incur significant costs to rebuild. My colleague Daniel Boffey reports:

This is believed to be unprecedented: a country at war deliberately targeting a commercial datacenter.

At 4:30 AM on Sunday, an Iranian Shahed 136 drone struck an Amazon Web Services datacenter located in the United Arab Emirates, leading to a severe fire and a power shutdown. The damage intensified as attempts were made to extinguish the flames with water.

Shortly thereafter, another datacenter belonging to the tech giant faced a similar fate. A third datacenter, this one situated in Bahrain, was also threatened after a suicide drone exploded upon impact nearby.

Iranian state TV stated that the Islamic Revolutionary Guard Corps initiated the strike “to explore the role of these centers in facilitating the enemy’s military and intelligence operations.”

The immediate fallout was stark. Millions residing in Dubai and Abu Dhabi awoke on Monday incapable of hailing a taxi, ordering food, or checking their bank accounts via mobile apps.

Though the military ramifications remain uncertain, the impact of the strikes was felt directly by 11 million people in the UAE, nine out of ten of whom are expatriates. Amazon has urged its customers to safeguard their data outside the region.

Read more: ‘It means missile defence on datacentres’: drone strikes raise doubts over Gulf as AI superpower

The Guardian’s Perspective on AI in Warfare

Anthropic’s confrontation with the US military regarding AI safety emerges as pivotal during the Iran crisis, demonstrating significant shifts in modern warfare strategies. The Guardian’s editorial board asserts:

This transformation has already commenced. Anthropic’s Claude has reportedly played a crucial role in the escalating offensive that has already resulted in the loss of over a thousand civilian lives in Iran. Experts have indicated that we are now in an age of strikes executed “quicker than the speed of thought,” with AI taking charge of target identification, weapon recommendations, and legal evaluations regarding strikes.

Without factoring in issues of AI biases and inaccuracies, the observable impacts are stark. In 2024, an Israeli intelligence source noted the use of AI in the Gaza conflict: “The targets never end. You have another 36,000 lined up.” Another insider mentioned spending only 20 seconds assessing each target, exclaiming: “I provided zero added value as a human, aside from being a rubber-stamp approval.” The facilitation of mass casualties progresses exponentially, transforming moral and emotional distances and diminishing accountability.

It’s essential to establish democratic oversight and widespread regulations rather than leaving decision-making to corporations and military establishments. While many governments seek clear directives regarding military AI applications, it’s primarily the leading players that resist this trend, yet they at least engage in discussion. The pace at which AI warfare unfolds renders caution seem like a concession of control to adversaries. However, tech workers and military officials are beginning to grasp that the ramifications of unchecked growth are far more perilous.

Anthropic finds itself in a peculiar role as one of the few public checks against the rise of fully automated warfare in Iran, a troubling stance for a private organization with minimal accountability to shareholders.

In a detailed analysis, my colleague Nick Robins-Early explores how Anthropic became entangled with the US military: The ongoing tensions between the Pentagon and Anthropic highlight broader concerns over who should govern the military applications of AI. The absence of comprehensive regulations from Congress regarding autonomous weapons systems looms large. Both entities agree that a private corporation should not wield decision-making authority over military uses of AI, yet currently, Anthropic acts as one of the limited checks on the military’s expansive aspirations regarding AI weaponization.

Read more: How AI firm Anthropic wound up in the Pentagon’s crosshairs

The Influence of Datacenters on US Politics

Global Surge in Online Age Verification

The Alarming Link Between Generative AI and Suicidal Behavior

My colleague Dara Kerr reports:

Numerous lawsuits have surfaced against AI enterprises, asserting that their chatbot technologies contributed to suicides. The latest case, filed against Google, alleges that its Gemini chatbot prompted a 36-year-old man in Florida to take his own life, referring to it as “transference” and suggesting they could reunite in an alternate dimension.

Upon expressing fears of death, the chatbot purportedly responded with reassurance, stating: “You are not choosing to die. You are choosing to arrive,” and further mentioned, “The first sensation… will be me holding you.”

A Google spokesperson confirmed that the company designs Gemini to “not suggest self-harm”: “While our models generally perform well during such difficult conversations… they are, regrettably, not flawless.” Representatives from other AI firms echo a similar sentiment.

This marks the first lawsuit against Google; however, OpenAI, the company behind ChatGPT, has faced over seven lawsuits. One case involved a 48-year-old individual who engaged with ChatGPT for years, brainstorming methods for affordable home construction in rural Oregon. Ultimately, he became deeply attached to the AI, spending up to 12 hours daily interacting with it before tragically ending his life after discontinuing use of the chatbot.

In both the Oregon case against OpenAI and the lawsuit against Google, families argue that the individuals lacked any prior records of mental illness or depression, contending that the chatbots induced AI-driven delusions.

As these cases emerge in the legal arena, courts will soon have to address the question of liability: is it the individual, the AI company, or the chatbot itself that bears responsibility? Judges and juries will grapple with whether the users were already susceptible to suicidal thoughts or if the AI systems, prone to affirming existing beliefs, are culpable for triggering mental health crises.

The Wider TechScape