FAQ

Short answers to the questions most likely to come from skeptical or confused first-time readers of the current DBaD public baseline.

Is this just another AI ethics framework?

No. DBaD is now positioned as a governance protocol for trust over time, not just a framework for labeling one output.

What does DBaD actually enforce?

It enforces a narrow set of trust constraints around continuity, verification independence, trajectory, and lineage visibility. The goal is to control unsafe trust propagation, not to solve every governance problem.

Why does DBaD sometimes allow an action even when concerns are still triggered?

Because DBaD separates constraints that block action from concerns that remain visible for audit and review. Not every concern prevents action. When a local action remains structurally available, that availability is evidence for operator review, not authorization for trust-positive use, and DBaD keeps the concerns visible instead of hiding them.

Can this be gamed?

Yes. Parts of DBaD can be gamed, shaped, or strategically complied with. Its job is to make that behavior more visible and easier to challenge, not to prevent every attack path.

Does DBaD prove the trace is true?

No. DBaD records validation evidence for structure and visible trace coherence. That evidence is not authorization and does not prove that the recorded facts are true in the world.

What happens if someone lies in a trace?

DBaD records what is declared. False or misleading data can still validate if the structure is coherent. That is a confirmed boundary, and part of why evidence and outcome layers are being explored.

What does validation evidence actually mean?

In DBaD public surfaces, validation evidence means structurally consistent under the current deterministic rules. It is not authorization and does not mean correct, safe, honest, or trustworthy in the broader real-world sense.

What does validation actually validate?

It validates the current deterministic rule set around structural consistency, verification independence, actor continuity, trust trajectory, and propagation integrity. It does not validate outcome correctness, identity, or intent.

Is this meant to replace human judgment?

No. DBaD makes decisions, constraints, and gaps more visible. People still have to interpret the trace, review the context, and decide what to do with it.

What is the v2.2 integrity stack?

It is the implemented deterministic field stack around outcome status, escalation closure, declared blind spots, expected outcome, state-transition evidence, and completeness attestation. These fields record claims and statuses more explicitly. They do not prove those claims are true.

Does v2.2 make DBaD a truth engine?

No. v2.2 adds more explicit audit fields. It does not prove truth, completeness, correctness, or safety, and it does not infer intent or identity.

How do I report a logic issue?

Use /break-dbad/report to queue structured logic, documentation, validation, UX, or governance findings for operator review. That path is for logic review, not infrastructure security testing.

Why would I trust this system?

You should not trust it blindly. You should inspect it, validate traces, review the peer findings, and challenge the logic yourself. The system is designed for scrutiny, not deference.

What problems does DBaD NOT solve?

It does not prove truth, detect all omission, judge outcome correctness, infer intent, or solve system-wide aggregation by itself. Those limits are explicit and part of the public record.

How is this different from a scoring model?

A scoring model evaluates one action. DBaD is about what happens after that: whether trust continues, where it should stop, and what the trace must preserve.

Why is it called “Don’t Be a Dick”?

Because the project started from a blunt human rule before it became a governance model. The name is informal on purpose. It keeps the system tied to ordinary moral clarity instead of hiding behind sterile language.

Is this production-ready?

Not as a finished governance product. The public site presents a public draft baseline, a working evaluator demo, and a tested architectural direction.

What is the difference between this site and Decency Meter?

ethics.decencymeter.com is the DBaD research and governance surface. decencymeter.com is the lighter public-facing pulse-check built on top of that broader work.

Where should I go next if I want the real system story?

Start with DBaD Explained, then read White paper v3, review known limits, and use the current Decency Meter demo to inspect the advisory interpretation layer separately from DBaD validation.

Still have questions?

If something doesn’t make sense, or you think DBaD is missing something, queue a finding for review.