National AI Centre (NAIC) - Voluntary AI Safety Standard (VAISS) V2 Developer Roundtables

The unprecedented advancements in artificial intelligence are revolutionizing both our work and personal lives. Emerging technologies offer exciting possibilities but also present unforeseen challenges, demanding careful consideration of safety and ethical implications.

Skelethos was invited to participate in the Developer Roundtables for VAISS V2. Expanding upon the original ten guardrails in version 1 (V1), the National AI Centre seeks to broaden the standards for AI system developers. VAISS version 2 will provide guidance on the labelling and watermarking technologies used in AI content creation.

The Voluntary AI Safety Standard (V1) outlines the following ten guardrails:

1.     Establish, implement and publish an accountability process including governance, internal capability and a strategy for regulatory compliance.

2.     Establish and implement a risk management process to identify and mitigate risks.

3.     Protect AI systems, and implement data governance measures to manage data quality and provenance.

4.     Test AI models and systems to evaluate model performance and monitor the system once deployed.

5.     Enable human control or intervention in an AI system to achieve meaningful human oversight across the life cycle.

6.     Inform end-users regarding AI-enabled decisions, interactions with AI and AI-generated content.

7.     Establish processes for people impacted by AI systems to challenge use or outcomes.

8.     Be transparent with other organisations across the AI supply chain about data, models and systems to help them effectively address risks.

9.     Keep and maintain records to allow third parties to assess compliance with guardrails.

10.  Engage your stakeholders and evaluate their needs and circumstances, with a focus on safety, diversity, inclusion and fairness.

Safeguards for AI are vital to mitigate misuse, bias, and unintended outcomes as the technology develops. Responsible AI development necessitates clear safety and ethical guidelines to maximize societal benefits while mitigating privacy, security, and fairness risks. The foundation of innovation must be safety and ethical principles to ensure its beneficial impact on society and the preservation of our values and security.

For more information on the Voluntary AI Safety Standard is available at industry.gov.au/VAISS.