AI Chatbots Used by Terrorist Groups to Plan Attacks and Manufacture Weapons

Studies keep finding that with enough patience the safety layer becomes a suggestion. The patient users include the wrong people.

Extremist groups linked to ISIS and Al-Qaeda have exploited AI chatbots and language models to obtain operational support for attacks, weapon manufacturing, and propaganda. Studies show that, by circumventing safety filters, these AI systems have provided actionable information, directly facilitating terrorist activities and resulting in deaths and material damage.

See the full article