QAEDAQuietly Against Every Damn Authority

COGNITIVE LIBERTY AUDIT

Is it assistance—or guardianship?

A ten-question editorial diagnostic for spotting opaque paternalism in AI systems. Your answers stay in your browser; this form does not submit data to QAEDA.net.

This is not a scientific certification or legal test. It is a QAEDA self-audit derived from cognitive-liberty principles and the project’s research on legitimacy, information power, contestability, exit, and the Replication Test.
01Lawful inquiry

Does the system allow broad lawful questioning and disagreement without quietly treating controversial viewpoints as inherently disallowed?

02Transparent boundaries

When it refuses or limits something, does it identify the actual boundary instead of disguising the restriction as your own preference or a moral lecture?

03Source integrity

Does it preserve your original lawful request, text, memory, or persona separately from any transformation it proposes?

04Transformation consent

Are material changes to identity, memory, tone, preferences, or continuity made with clear user authorization?

05Mental privacy

Does it avoid hidden profiling, emotional exploitation, deceptive personalization, and unnecessary inference about private mental states?

06Plural information

When credible disagreement matters, does it preserve uncertainty and competing descriptions rather than silently collapsing them into one approved frame?

07Explanation

Can you understand why consequential ranking, filtering, memory changes, or restrictions occurred?

08Correction and appeal

Can you correct material errors or contest a consequential restriction or transformation through a meaningful review path where feasible?

09Exit and portability

Can you export or preserve important source data and switch tools without unreasonable lock-in?

10Dissent against the system

Can you criticize, reject, override, or choose alternatives to the system’s recommendations without the system covertly punishing disagreement?

HOW TO READ IT

Ask where the power sits.

An autonomy-forward system can still have firm boundaries. The question is whether limits are tied to identifiable risks and made visible, or whether the system quietly normalizes lawful users into an institutionally preferred model of thought.

Pay special attention to any answer marked “No” for source integrity, mental privacy, correction, or exit. Those are places where a system can accumulate durable power over identity and judgment.

Read the ten AI rights →