Security News > 2023 > November > Released: AI security guidelines backed by 18 countries
The UK National Cyber Security Centre has published new guidelines that can help developers and providers of AI-powered systems "Build AI systems that function as intended, are available when needed, and work without revealing sensitive data to unauthorised parties."
The Guidelines for secure AI system development cover four key stages of the development lifecycle of machine learning applications.
Secure design hinges on all persons involved - system owners, developers, users - being aware of the unique security risks facing AI systems and being taught to avoid them.
"Model the threats to your system, and design your system for security as well as functionality and performance," the guidelines instruct.
"The new UK-led guidelines are the first of their kind to be agreed globally. They will help developers of any systems that use AI make informed cyber security decisions at every stage of the development process - whether those systems have been created from scratch or built on top of tools and service provided by others," the NCSC pointed out.
We urge all stakeholders to read these guidelines to help them make informed decisions about the design, deployment and operation of their machine learning AI systems.
News URL
https://www.helpnetsecurity.com/2023/11/27/ai-security-guidelines/
Related news
- U.S. Government Releases New AI Security Guidelines for Critical Infrastructure (source)
- AI and the future of corporate security (source)
- In the rush to build AI apps, please, please don't leave security behind (source)
- Generative AI Security - Secure Your Business in a World Powered by LLMs (source)
- GitHub Launches AI-Powered Autofix Tool to Assist Devs in Patching Security Flaws (source)
- Enterprises increasingly block AI transactions over security concerns (source)
- Google Cloud/Cloud Security Alliance Report: IT and Security Pros Are ‘Cautiously Optimistic’ About AI (source)
- Six steps for security and compliance in AI-enabled low-code/no-code development (source)
- Security pros are cautiously optimistic about AI (source)
- 92% of enterprises unprepared for AI security challenges (source)