The rapid advancement of Artificial Intelligence (AI) has opened up a world of possibilities in various fields, including medicine and pharmaceuticals. AI-based drug discovery has already proven its worth by helping scientists find new forms of antibiotics and medication to combat Covid-19.
However, the same technology that has the potential to save lives could also be easily misused for malicious purposes, such as the design of highly toxic chemical and biological weapons, warn researchers.
A team of four researchers involved in AI-based drug discovery was asked by the Swiss Federal Institute for Nuclear, Biological, and Chemical Protection to examine the potential misuse of Artificial Intelligence to design biological weapons. Their AI came up with 40,000 potentially toxic drugs in a six-hour experiment.
The results, published in the journal Nature Machine Intelligence, served as a wake-up call for the researchers and a warning of the dangers of misusing Artificial Intelligence to design biological weapons.
Lead author Fabio Urbina of Collaboration Pharmaceuticals in North Carolina, USA, stated that the experiment made them realize “all we have to do is flip the switch round and say, ‘instead of going away from toxicity, what if we go towards toxicity?’” The AI algorithm started “producing all of these toxic molecules, a lot of which look like chemical warfare agents,” Urbina said. Some of the molecules produced were even “more toxic than VX,” a nerve agent developed in the 1950s by the UK’s Defence Science and Technology Lab, which causes death by muscle paralysis.
The ease of the experiment was the most concerning aspect, as many of the resources used were readily available for free online. This highlights the potential for individuals with malicious intent to misuse Artificial Intelligence to design biological weapons.
While AI can revolutionize various fields, including medicine and pharmaceuticals, it’s crucial to be aware of its potential dangers. The experiment conducted by the Swiss Federal Institute serves as a warning that AI algorithms could easily be manipulated to search for toxic nerve agents and underscores the importance of being cautious when using Artificial Intelligence to design biological weapons.
