AI Technology Heightens Nuclear War Risk, Warns UN Chief

UN Secretary-General António Guterres has raised a stark warning about the increased threat of nuclear warfare due to the rapid expansion of artificial intelligence (AI) technology. Speaking through a recorded video at the annual meeting of the US Arms Control Association (ACA) in Washington, Guterres highlighted the critical juncture humanity faces as technological advancements in AI amplify the risks associated with nuclear weapons.

The Weakening of Non-Proliferation Systems

Guterres expressed deep concern over the weakening systems designed to prevent the use, testing, and proliferation of nuclear weapons since the end of the Cold War. “The regime designed to prevent the use, testing, and proliferation of nuclear weapons is weakening,” he stated, with urgency heightened as the New START accord between the US and Russia nears its expiration in 2026.

The Role of AI in Nuclear Launch Procedures

There is a growing fear that nations, particularly the US and Russia, might employ AI to streamline nuclear launch procedures, potentially bypassing human oversight. Both countries maintain their intercontinental ballistic missiles on high alert, ready to launch at a few minutes’ notice. Guterres underscored the necessity for human control over nuclear decisions, declaring, “All countries must agree that any decision on nuclear use is made by humans, not machines or algorithms.”

Call for International Agreements

In 2022, the US, UK, and France issued a joint statement to maintain human control over nuclear launches. However, Russia and China have yet to follow suit. Guterres emphasised the need for nuclear states to renew their commitment to disarmament and non-proliferation. He urged the US and Russia to resume dialogue, fully implement the New START treaty, and negotiate its successor. “They must resume dialogue and commit to preventing any use of a nuclear weapon,” he insisted.

Declining but Still Dangerous Nuclear Stockpiles

According to the Federation of American Scientists, the number of nuclear weapons has dramatically decreased from a peak of about 70,300 in 1986 to an estimated 12,100 this year. Despite this reduction, the total military stockpile of warheads available for use has begun to increase again after years of decline.

The Ethical Debate Over AI in Military Systems

The integration of AI into military systems, including nuclear command and control, has sparked ethical debates. There are significant concerns about the potential for AI to misinterpret data or make erroneous decisions, which could lead to unintended escalation. The rapid pace of AI development could also outstrip international policy efforts, complicating disarmament and non-proliferation.

Scientists’ Warnings and the Path Forward

Guterres’ warnings echo those of many scientists who have long cautioned about the dangers of AI. A 2018 Rand Corporation report highlighted how AI could increase the risk of nuclear war by improving military capabilities to track and destroy a rival’s nuclear forces. Prominent figures like former US Secretary of State Henry Kissinger have called for international cooperation to manage AI risks.

Advancing Disarmament Efforts

The Treaty on the Prohibition of Nuclear Weapons (TPNW), which came into force in January 2021, represents a significant step towards legally binding nuclear disarmament. AI also holds potential for non-weapon applications, such as enhancing treaty verification through advanced data analysis.

Conclusion

The call to action is clear: nuclear-armed states must engage in earnest dialogue and take concrete steps to ensure that AI does not undermine global security. The balance of power and the safety of humanity hinge on maintaining strict human control over nuclear arsenals and committing to comprehensive disarmament.