AI isn’t a future threat—it is already erasing Gaza

Much is being said right now about the dangers posed by AI and the possibility that it could ultimately wipe out humanity. Some exp erts, such as Evan Hubinger, Anthropic’s Alignment Science Lead and one of the AI industry’s leading safety researchers, have gone much further, warning publicly that there is more than a 10 percent chance that AI could kill all humans within the next decade.

When senior executives from some of the world’s leading AI companies met in San Francisco for Salesforce’s Dreamforce conference, held 15th–17th September 2026, they differed sharply over the seriousness of the risks posed by increasingly powerful AI and over how those risks should be managed. They did, however, share a fundamental belief: that the industry itself is capable of developing safeguards, testing systems, monitoring their behaviour and preventing or correcting dangerous failures. Where they differed was over how much restraint was necessary, who should impose it, and whether governments should intervene through new legislation and regulation.

The leading AI developers’ position is aligned with that of President Donald Trump, He wants AI to realise its full potential as quickly as possible and supports the industry’s drive to keep the AI race moving. Trump’s position is essentially this: let AI run; give developers the time, resources and freedom they need to take the technology to the next level. Trump, whose country is currently leading the AI race, has dismissed warnings about AI’s potential to destroy humanity as a “HOAX,” just as he has dismissed warnings about climate change despite the scientific evidence. He has also repeatedly argued that the US cannot afford to slow the development of AI while China is racing on. Just yesterday, before the United Nations he said: “I’m not going to stifle growth of something that will be bigger than the Industrial Revolution.” 

To Trump, AI is another strategic opportunity that his country cannot miss: develop it, exploit its economic and strategic potential, and make sure China stays behind. In that sense, it fits a broader strategic logic in which Washington seeks to secure control or advantage over resources and technologies considered vital to American power, rather than leave them available to rivals.

Simply put, the logic sounds like this: let the flock of sheep spend the night with a pack of hungry wolves, and see what happens in the morning.

What is striking about the latest discussion among the leaders of the AI industry in San Francisco is that, while some of them referred to the possibility of AI causing catastrophic harm to humanity, few, if any, pointed to a specific, real-world example of what that danger could already look like. More striking still, no one used Gaza as the obvious real-world example: how AI has already been incorporated into warfare and how technology supplied by some of their own companies has been used by the Israeli military in its war on Gaza.

Some of those present know, from the experience of their own companies, that their technologies have been used extensively in support of the Israeli military.

Commercial AI models from both Microsoft and OpenAI have been deployed by the Israeli military, along with Microsoft’s Azure cloud services, while Google and Amazon have provided cloud computing and other AI services to the Israeli occupation military under Project Nimbus.

The Associated Press has documented how these technologies have been used to process intelligence, translate and analyse intercepted communications, and support the identification of targets in Gaza and Lebanon. The question, therefore, is not whether AI could one day become a danger to humanity. In Gaza, the question is what it is already doing when placed in the hands of a military at war.

READ: Board of Peace to unveil $2.45B Gaza recovery plan: Report

The human consequences are before the world. More than 75,000 Palestinians have been killed since October 2023, while the destruction of housing alone has affected some 1.7 million people, with more than 320,000 homes damaged. The United Nations estimates that 1.98 million people—94 percent of Gaza’s population—required shelter assistance in its latest assessment. Gaza’s infrastructure has been devastated, while essential services, including health, water and sanitation, remain severely compromised.

This is not a hypothetical future in which an uncontrolled machine decides to destroy humanity. It is a real war, taking place now, in which AI is increasingly being incorporated into systems of surveillance, intelligence and targeting. Yet no one at the summit appears to have asked the obvious question: if these companies know what their AI is doing in Gaza, why is Gaza not the first example they cite when discussing the dangers of this technology?

And if AI can already contribute to this scale of destruction in Gaza, what happens when the technology becomes even more powerful, more autonomous and more deeply embedded in warfare elsewhere in the world?

In addition to this, there is now documented evidence that Israel has sought to interfere with the information on which AI systems train, by flooding the online environment with misleading and fabricated material designed specifically to influence what AI models tell users. Israel funded the creation of a fake American think tank called the Hanover Institute for Public Policy, whose stated purpose appeared to be legitimate research but which had no apparent legal existence, physical address, named researchers or bylines on its reports. Created on behalf of Israel’s Government Advertising Agency through the US public-relations firm Piro Inc., the institute produced more than 560,000 words across 124 reports in just nine days, much of it focused on Israel, Palestine and Gaza. The material was deliberately structured and optimised for discovery and citation by AI systems, effectively attempting to make propaganda look like credible research and feed it into the information environment from which AI systems generate answers.

This goes far beyond the conventional hacking of communications or the gathering of intelligence. It represents an attempt to manipulate the information environment itself—to establish databases and bodies of supposedly authoritative knowledge built partly on fabrication, distortion and selective framing, so that AI systems encounter and potentially reproduce those claims as they become increasingly capable of simulating human reasoning. In other words, Israel is not only using AI in the war in Gaza for surveillance, intelligence, communications and the identification of targets; it is also seeking to interfere with the informational foundations on which AI systems operate, attempting to shape what they learn, what they retrieve and ultimately what they tell human beings.

OPINION: The Mangoush affair, three years on: Palestine remains a red line in Libya’s foreign policy

The views expressed in this article belong to the author and do not necessarily reflect the editorial policy of Middle East Monitor.