AI Safety
What is AI Safety?
Research and practices aimed at ensuring AI systems operate safely, reliably, and without causing unintended harm to individuals or society.
Terms that appear alongside ai safety
Each of these is named in at least one of the same controls as ai safety. The number is how many controls name both.
- governance 11 shared controls
- nist 9 shared controls
- ai governance 8 shared controls
- cybersecurity 8 shared controls
- transparency 7 shared controls
- eu ai act 7 shared controls
- governance framework 6 shared controls
- compliance 5 shared controls
Frameworks that govern ai safety
What the standards actually require on ai safety
Requirements naming ai safety across 6 standards, quoted from the control text.
Escalate concerns about non-compliance with the policy, AI incidents or broader AI safety issues through the agency's governance and escalation processes.
nsw-ai-operational-policy-2026::5.6.c · 5.6 c Report issues →By 15 December 2026 every agency needs a documented way of embedding responsible AI practice, sized to how much AI it uses. Five elements are the floor.
australia-policy-for-the-responsible-use-of-ai-in-government-version-2-0::3.1 · Establish an approach that operationalises the responsible use of AI within twelve months →Article 6 designates the high-risk machinery categories listed in Annex I that are subject to mandatory third-party conformity assessment (because of the level of risk).
MR-Art.6_7 · High-risk machinery categories and safety components (Articles 6-7 + Annex I) →FIRST standards pipeline + 2024-2025 roadmap. CURRENT STANDARDS: CSIRT Services Framework v2.1 (2019) + CVSS v4.0 (2023) + TLP v2.0 (2022) + IEP v2.0 + MPCVD Guidelines + PSIRT Services Framework v1.1 (2020).
FIRST-Status-Pipeline · FIRST Standards Pipeline - 2024-2025 Roadmap and Coordination with NIS2, CIRCIA, EU CRA →ISMAP Assessment positions ISMAP within the comprehensive Japanese and international cloud security regulatory landscape. (1) External Assessment by ISMAP-Approved Auditor: CSP must undergo annual third-party assessment by an audit organisation on the ISMAP Pr...
ISMAP-Assessment-ExternalAuditor-AnnualReview-CustomerTransparency-Coord-FedRAMP-IRAP-GCloud-PIPA-ISO27017 · ISMAP Assessment - External ISMAP-Approved Auditor + Annual Review + Customer Information and Transparency + Coordination FedRAMP/UK G-Cloud/Australia IRAP/Singapore MTCS + ISO 27017 + PIPA + Japan Digital Agency →Prioritise research to mitigate societal, safety and security risks and prioritise investment in effective mitigation measures. Code of Conduct detail: conduct, collaborate on and invest in research that advances AI safety, security and trust, including uphold...
japan-ai-guidelines::HP.VIII · D-VIII. Research and investment in risk mitigation →Questions people ask about ai safety
What is AI Safety?
Why is AI Safety important for compliance?
Which compliance frameworks address AI Safety?
Where can I learn more about AI Safety?
See how AI Safety applies across compliance frameworks
Our platform maps 1,016 frameworks with 316K cross-framework control mappings. Explore how this concept is addressed across standards.