They’ll use new tech to perform spoiler attacks – extinction.

A new paper published in the European Journal of Risk Regulation considers the danger from existential terrorism, defined as acts that threaten the existence of humanity. The authors highlight what they term ‘spoiler attacks’ involving AI or other new technology, which might enable a group with limited resources to cause unprecedented destruction.

“I don't expect existential terrorism to be at the top of global agendas, nor do I believe it should be,” Zachary Kallenborn, one of the report’s, authors told me. “But global discourse is clearly changing around existential risk.”

Kallenborn is a Policy Fellow at the Schar School of Policy and Government, an officially proclaimed U.S. Army 'Mad Scientist', and a national security consultant. The paper is part of a special issue on long-term risks and special governance, with the unexpected effects of emerging technology being a key consideration.

“Technology is definitely bringing more power to the people,” says Kallenborn. “The open question is how much capability is really needed to generate existential harm.”

Kallenborn notes that unlike state actors, terrorist groups generally lack capacity to build effective weapons of mass destruction such as nuclear warheads. The best-known apocalyptic group, the Aum Shinrikyo cult, carried out several research projects including work on biological warfare. But they were forced to scale back their ambitions, and the cult's final effort was a nerve gas attack on the Tokyo subway in 1995 which caused fourteen deaths and affected thousands more. This was an appalling total, but still far short of the group’s apocalyptic goal.

Rather than developing a superweapon themselves, a modern terrorist group could carry out a form of sabotage, a spoiler attack, to cause a cataclysm.

For example, terrorists could leverage the potential risks in advanced AI research, an area which which some warn carries “risk of extinction,” and leading to calls for strict safeguards on research. Rather than building their own super-intelligent AI, terrorists might carry out a spoiler attack to break through the safeguards preventing an AI from being developed beyond a certain stage or released. This might be carried out remotely via hacking, on the spot by recruiting or subverting researchers, or by an armed intrusion into a research facility.

Spoiler attacks might also target biological research or nanotechnology projects, both areas where high levels of safeguarding are required. The authors note that new tools such as CRISPR, rapid DNA sequencing and DNA/RNA synthesis mean that there are now far more groups working on potentially hazardous biological projects. The unproven lab leak theory that COVID-19 escaped from a Chinese research facility could be a blueprint for a spoiler attack.