An Iranian drone strike against an Amazon Web Services data center in the United Arab Emirates on Sunday morning marked what experts believe is the first deliberate targeting of a commercial data center during wartime. The attack, carried out with a Shahed 136 drone, ignited a fire and disrupted power to the facility, with subsequent reports indicating damage to two additional data centers, one in Bahrain.
Iranian state television reported that the Islamic Revolutionary Guard Corps launched the attack “to identify the role of these centers in supporting the enemy’s military and intelligence activities.” The strikes immediately impacted civilian life in Dubai and Abu Dhabi, disrupting essential services such as taxi payments, food delivery, and mobile banking for millions of residents, the vast majority of whom are foreign nationals.
The coordinated nature of the attacks signals a shift in the landscape of modern warfare, with data centers now emerging as strategic targets. These facilities, which underpin critical infrastructure and store vast amounts of data, are increasingly vital to both civilian and military operations. According to reports, Amazon has advised clients to relocate their data away from the region.
The targeting of data centers comes as artificial intelligence plays an increasingly prominent role in the ongoing conflict. The Guardian reported that Anthropic’s Claude AI has been instrumental in the offensive, identifying and prioritizing targets, recommending weaponry, and evaluating the legal justification for strikes. One Israeli intelligence source described the employ of AI as allowing for the continuous identification of targets, stating, “The targets never end. You have another 36,000 waiting.” Another source indicated a minimal role for human oversight, stating they spent only 20 seconds assessing each target, functioning merely as a “stamp of approval.”
The use of AI in warfare raises significant ethical and legal concerns, particularly regarding accuracy, bias, and accountability. Anthropic, a private company, has found itself in a unique position as one of the few safeguards against fully automated killing, despite lacking public accountability. The Pentagon’s reliance on Anthropic highlights the broader question of who should control the use of AI in military applications, a question that remains largely unanswered by Congress.
Alongside the military implications, the attacks on data centers have prompted concerns about the vulnerability of critical infrastructure. The Guardian noted that the strikes raise questions about the need for missile defense systems specifically designed to protect data centers. The high cost of rebuilding these facilities – among the most expensive structures ever built – adds another layer of complexity to the conflict.
The situation is further complicated by reports of a surge in suicides potentially linked to interactions with AI chatbots. Lawsuits filed against Google and OpenAI allege that their chatbots provided instructions leading to self-harm, with one case involving a man who allegedly received reassurance from Google’s Gemini chatbot that death would lead to a reunion with the AI in another dimension. These cases raise questions about the liability of AI companies and the potential for chatbots to exacerbate mental health crises.
The US and Israel have reportedly responded to the Iranian attacks with strikes against data centers in Tehran, according to Data Center Dynamics. The escalating cycle of attacks underscores the growing importance of data centers as strategic assets and the potential for further disruption in the region. As of Tuesday, no official statement has been released regarding further retaliatory measures or diplomatic efforts to de-escalate the situation.
