AI will help us improve cyber defence over time; however, it also accelerates the speed, scale and sophistication of cyber threats. These technologies can generate text, code, images and video, and automate repetitive or complex tasks to improve efficiency and productivity. However, as Australian organisations integrate AI and large language models into their environments, new and potentially exploitable security risks are emerging, expanding their overall opportunities – or our attack surface.
Over the last financial year, ASD’s ACSC has observed the increased risks and exposure that advances in AI models have introduced into Australian organisations.
