CSAEAIIndependent observatory
INSTITUTION

OUR POSITION

An independent
standard.

CSAEAI examines how artificial intelligence systems can become more understandable, assessable and accountable.

Start a conversation

MISSION

Audit as a discipline of evidence, not a promise.

Our work begins with a simple observation: as models gain capability and autonomy, their evaluation must extend beyond statements of intent.

We bring together technical inquiry, the context of use, legal requirements and human consequences. We publish the questions, references and conclusions that can inform public debate.

IndependenceExamine evidence without confusing reference with endorsement.
TraceabilityConnect every public position to the evidence on which it rests.
ProportionalityMatch the depth of scrutiny to the risks and contexts of use.

DISCLOSURE DOCTRINE

What we disclose.
What we protect.

PUBLICVisible · Citable · Open to challenge

Mission, research scope, reference frameworks, analysis, limitations and editorial positions.

PROTECTEDRestricted · Compartmentalised · Not published

Internal procedures, assessment instruments, confidential data, decision rules and security safeguards.

REFERENCES

Frameworks cited with precision.

These references inform our analysis according to context. Their inclusion does not imply certification, partnership or endorsement of CSAEAI by the organisations cited.

PUBLIC REFERENCE
LAW

European Union

The AI Act, GDPR and guidance from the relevant authorities

PUBLIC REFERENCE
STANDARDS

Standardisation

ISO/IEC 42001 and other relevant standards

PUBLIC REFERENCE
RISK

Risk management

NIST AI Risk Management Framework

PUBLIC REFERENCE
ETHICS

International principles

OECD principles and the UNESCO Recommendation on the Ethics of Artificial Intelligence

PUBLIC REFERENCE
Notre position | CSAEAI