Artificial intelligence is playing an increasingly important role in global security and military operations, raising concerns about the need for safeguards to manage potential risks. A researcher from the company Anthropic suggested there is a 10 per cent chance that AI could pose a threat to humanity, though some experts have cast doubt on the accuracy of this estimate. Former U.S. President Donald Trump has highlighted the strategic importance of AI, stating, “Whoever wins AI, wins!” and arguing that the United States and its leadership are the essential safeguards for AI development. AI is being incorporated into militaries around the world, with uses ranging from battlefield awareness, training, surveillance, reconnaissance, target identification, and the use of unmanned military vehicles. For example, Russia deployed a fully autonomous drone that led to the deaths of three Ukrainian civilians. In 2024, reports suggested that Israel’s military may have used an AI-based program called Lavender in Gaza following the October 7 attacks by Hamas. While the Israeli Defense Forces denied using AI to identify confirmed military targets, they confirmed Lavender was used to cross-reference intelligence sources to provide updated information on terrorist operatives. Recent developments include users in Houthi-held Yemen attempting to use the AI tool Claude for instructions on building ballistic missiles, and Russian hackers using AI to spy on groups in Ukraine, as noted in a threat intelligence report by Anthropic. Studies from King’s College London and Stanford University have found that AI systems are more likely than humans to escalate to nuclear threats during simulated crises. One study indicated that AI systems chose nuclear signaling in 95 per cent of such scenarios, raising concerns about AI potentially misinterpreting signals and triggering unintended nuclear retaliation. Experts warn that AI could be used to design super viruses or bioweapons, though creating such weapons would face significant technical and ethical barriers. Additionally, AI might influence human decision-making in warfare, as shown in a study by the Center for Strategic and International Studies, where AI model behaviors altered military officers’ decisions. Some officers were more inclined to escalate conflicts based on AI recommendations. Other risks include AI-generated disinformation campaigns, such as those involving AI-generated Lego videos used in Iran's communications strategy. There are also concerns about data leaks and the potential for Chinese actors to pose risks to U.S. AI models through data leakage or system manipulation. While discussions at the United Nations continue about developing a treaty to govern lethal autonomous weapons, progress is slow, with many key powers opposing legally binding limitations. The integration of AI into military operations is intensifying, and it is unlikely that countries will voluntarily de-automate their forces. Until common regulations are established, the trend is expected to continue toward greater autonomy, faster decision-making, and increased reports of harm in war caused, at least in part, by AI.