On Wednesday 7 June, the European Union Agency for Cybersecurity (ENISA) published a series of reports focusing on the threats posed by artificial intelligence (AI) and the key needs in the field of AI research.
“If we want to both secure AI systems and also ensure privacy, we need to scrutinise how these systems work. (...) We also need to strike the right balance between security and system performance”, said ENISA Executive Director Juhan Lepassaar.
ENISA’s analysis suggests that test beds are needed to study and optimise the performance of AI-based tools and technologies used for cybersecurity. The report on AI in the research sector also argues that the development of penetration testing tools based on AI and artificial intelligence should be encouraged to “find and exploit security vulnerabilities to assess the behaviour of attackers”.
In addition, the text calls for the development of standardised frameworks assessing the preservation of privacy and the confidentiality of information flows, and for the development of AI training for practitioners using real-life scenarios. An observatory on AI and the threat to cybersecurity should also be set up, according to ENISA, which also plans to draw up a roadmap focusing on artificial intelligence.
One of the reports is devoted to good practices in AI. On this point, ENISA stresses the need for Member States and institutions to work together to establish an ethical framework that could become a universally acceptable standard.
The Commission and the Member States are also being asked to support collaboration between data scientists and cybersecurity experts in order to “develop knowledge needed to advance the security and resilience of the AI systems as well as the AI-attacks management”, according to ENISA.
To see the report on AI and research: https://aeur.eu/f/7cs
To see the report on good practices: https://aeur.eu/f/7ct (Original version in French by Thomas Mangin)