CYBERTheatIntel: AI – UK & US Spooks publish AI Development Guidance
Tuesday, 28th November 2023
CYBER THREAT INTEL: AI – UK & US Spooks publish AI Development Guidance
Posted by: Iain Fraser – Cybersecurity Journalist Gibraltar
via IainFraser.net/ Daily Cyber Insights
CYBER THREAT INTEL: AI – UK & US Spooks publish AI Development Guidance
The UK NCSC & US CISA have joint-published what they term as “security-focused guidance” for AI & ML developers and those who leverage AI/ML with their protocols and systems.
The Publication of Guidelines for Secure AI System Development (PDF), apply to all types of AI/ML systems, regardless of whether built from wireframe up or added as a bolt-on on third-party resources, to address issues related to AI, cybersecurity, and critical infrastructure.
Devised to be used in conjunction with existing Cybersecurity, Incident Response and Cyber Risk-Management protocols. The NCSC and CISA have said “Providers should implement security controls and mitigations where possible within their models, pipelines and/or systems, and where settings are used, implement the most secure option as default”
The Guidelines have been collated in collaboration with over 20 domestic and international cybersecurity organisations, broken down into four sections, covering different stages of the AI system development, namely – Design, Development, Deployment & Operation and Maintenance. Implementation of the recommendations will require investments in features, mechanisms and tools designed and implemented to protect customer data at all layers and throughout the entire system lifecycle.
