The 5 Pillars of Evidentiary Accountability
When a machine-led outcome triggers a crisis, traditional system logs leave massive structural gaps. The DAREB test applies a brutal, binary standard to your records, verifying whether five distinct links survive intense scrutiny:
- Decision: What exact automated outcome occurred, who or what triggered it, and who is legally responsible for its consequences?
- Authority: Where is the unbroken legal and operational chain that gave the automated system the right to make that specific determination?
- Record: Do you possess a permanent, immutable, and tamper-evident log of the execution state captured at the precise millisecond of runtime?
- Evidence: Can you produce the exact, uncorrupted inputs, context parameters, and prompt states that the system ingested to produce that result?
- Basis: Can you explicitly demonstrate the underlying logic, model weights, or ruleset that connected the evidence to the final decision?
If any link in this sequence is missing, fragmented, or opaque, the chain of responsibility is broken. The decision cannot stand.
The Director Accountability Test
Oversight habits designed for a slower world fail when automated systems operate at scale. When a failure triggers a crisis, investigators and D&O insurers look directly past corporate entities to examine individual board members.This test uses five personal markers: Participation, Information, Understanding, Judgement and Evidence to determine if your individual actions stand up under hostile cross-examination.
The EU PLD Exposure Test.
The revised Product Liability Directive brings strict liability rules straight to software, AI systems, and automated updates. If your system causes harm and a court orders an evidence disclosure, gaps in your technical logs create an automatic legal presumption that your AI was defective.This 16-question framework maps your exposure across five critical structural vectors before a trial begins.