AIDuty.orgv0.1.0-Alpha
AboutFrameworkCommunityApply Now

© 2026 AIDuty.org. All rights reserved. Defining the ADS-2026 Framework.

Draft v0.1.0-Alpha|
Governance & Legal
Draft v0.1.0-AlphaPublic Consultation Phase

ADS-2026 Framework

Standard for the Duty of Care in Artificial Intelligence

0 Contributors
Updated just now
1.0

Introduction & Scope

The ADS-2026 standard defines the mandatory Duty of Care requirements for entities developing, deploying, or operating Artificial Intelligence systems with significant societal impact.

2.0

Core Principle: Human Wellbeing

All AI systems must be optimized for the preservation and enhancement of human physical, psychological, and social wellbeing.

VR-2.1

Systems must undergo a mandatory "Human Impact Assessment" (HIA) prior to deployment.

0
VR-2.2

Optimization functions must explicitly exclude metrics that incentivize psychological harm or addiction.

0
3.0

Pillar I: Transparency

Transparency is the foundation of accountability. Entities must maintain verifiable records of system behavior and data provenance.

VR-3.1

Entities must maintain immutable "Inference Logs" for all high-stakes autonomous decisions.

0
VR-3.2

Training data lineage must be documented and accessible to accredited third-party auditors.

0
VR-3.3

Model limitations and known failure modes must be disclosed via standardized "ADS Model Cards".

0
4.0

Pillar II: Neutrality

AI systems must not be used to manipulate public discourse or amplify systemic biases.

VR-4.1

Algorithmic preference for specific political or commercial ideologies is strictly prohibited.

0
VR-4.2

Continuous bias monitoring and mitigation protocols must be active during the entire lifecycle.

0
5.0

Pillar III: Safety

Robust fail-safes and circuit breakers must be integrated into the core architecture of critical systems.

VR-5.1

Mandatory "Hardware Kill-Switch" for AI systems controlling physical infrastructure.

0
VR-5.2

Real-time "Out-of-Distribution" (OOD) detection must trigger immediate human-in-the-loop intervention.

0

Contents

1.0 Introduction & Scope2.0 Core Principle: Human Wellbeing3.0 Pillar I: Transparency4.0 Pillar II: Neutrality5.0 Pillar III: Safety

How to contribute

Click on any Verifiable Requirement (VR) to view community feedback or submit your own technical amendment.