Skip to content

Artificial intelligence and the cyberthreat landscape: understanding and navigating the challenges

Artificial intelligence (AI) is revolutionising business and industry as well as the everyday lives of individuals – and its take-up is growing. However, the more we rely on AI systems, the more these systems will be targeted by malicious actors.

How this happens is the subject of this whitepaper. Its focus is AI and cybersecurity in general and these three types of attacks: data poisoning, white-box evasion and membership inference attacks.

These cyberattacks and their effects on AI systems are examined in detail, along with an overview of the interventions that aim to enhance the cybersecurity of AI. It also focuses on the first countries and regions to address this issue: the USA, the UK, the EU, Germany and China, where measures to strengthen AI systems against cyberattacks have been discussed in several official publications, along with the responsible development of AI systems and general cybersecurity measures.