AI-developed poison gas: how real is the danger?

Collaborations Pharmaceuticals is just one small company in a universe of hundreds of companies that use artificial intelligence (AI) for, for example, drug discovery. The authors ask how many of them would have thought to manipulate or even abuse things with their abilities. With technological advancements such as AI, is the danger of new weapons making you sick, crippling entire body functions or even killing living things?

No reason to panic
There is no need to panic, says political scientist Frank Sauer of the Bundeswehr University in Munich. “But the risk is there.” The so-called dual-use problem is particularly pronounced in this area: that technologies or goods can be used for both civilian and military purposes.

Natural sciences are good at collecting data, says Alexander Schug of the Steinbuch Center for Computing at the Karlsruhe Institute of Technology. For example, there are well-prepared databases for drugs and protein structures, many of which are publicly available. “That makes them predestined for AI.” But even a good programmer should find his way around the structure of the datasets, Schug says. “It’s all on a very theoretical level at the moment.” In addition to sufficient training, sufficient computer capacity is also required. “You can do great calculations with graphics cards, but not really big calculations.”

Schug admits: “It could be a new kind of danger if new toxins and new synthetic routes are developed on the computer.” But the big question is always whether the substances can be synthesized at all – ie manufactured. Some compounds may not hold, in other cases there is a lack of raw materials.

AI-designed molecules as poisonous gas
Potential uses would also depend on this: AI-engineered molecules as a poisonous gas attacking cells through the respiratory tract? Like an invisible substance that picks you up when you grab something? Or manipulated pathogens that are spread through drinking water systems?

International disarmament adviser Ralf Trapp speaks of a “whole chain of things” that are necessary before anything can be deployed militarily. The technical possibilities shortened the time until something came to the market – this was also reflected in the development of the corona vaccines. But to implement it as a weapon, more resources are needed. “What happens in the lab is one thing. What can come of it is something completely different,” Trapp says. “That doesn’t mean we don’t have to worry, it’s always a race against time.”

According to Una Jakob of the Hessian Foundation for Peace and Conflict Research, awareness of the potential danger is growing. In her view, however, research and companies could be better informed about the risks: “Not every scientist is aware that research can also be misused.” This should already be pointed out during the training. The more research that is done, the greater the danger that someone will work in a laboratory that has not only good intentions.

Eclipse makes discovery difficult
The potential eclipse makes detecting bioweapons programs difficult, says Jakob. Many substances and materials are used medicinally or pharmaceutically. It’s hard to weigh – whether inspectors should be able to look behind the scenes of the arms industry. Sauer of the Bundeswehr University explains that it is not easy to ban individual substances because of the double use – both positively and negatively.

Yet Jakob also estimates the risk of new bioweapons that have been developed at a high technological level as low: “I think it is unlikely that someone will do complicated genetic research if he plans an attack. Then he can also order ricin online.” That means less effort in time and money.

And even if the worst happens, technological advances can have one advantage, as Trapp says: the ability to react faster. Ideally, with the scientific capabilities, an antidote could be found more quickly – even if the trigger was developed with AI.