FOR COMPANIES
You use AI for decisions.
Be receipt-ready before August 2026.
Every important AI decision should come with a receipt.
If there is no receipt, there should be a clear way to request one.
Receipt or Request. Nothing in between.
EU AI ACT — DEADLINE: AUGUST 2, 2026
AI systems used in hiring are classified as high-risk under the EU AI Act. Documentation, human oversight, and transparency requirements are enforceable from August 2026. Penalties can reach €15 million or 3% of global revenue.
The law has extraterritorial reach — it can affect organizations outside the EU when AI-assisted hiring decisions affect people in Europe. Audrith helps teams become receipt-ready before this becomes a problem.
WHY RESPOND
Closes the case
A receipt or response marks the case as resolved. The Transparency Pending status disappears.
Demonstrates responsibility
Companies that respond — even with context — are publicly visible as engaging in good faith.
Creates a verifiable record
A receipt creates a visible record that can be checked and shared using a stable Record ID.
Avoids escalation
A non-response keeps the case public indefinitely and may attract media or legal attention.
HOW TO RESPOND
01
Receive the notification
You receive an email with a secure response link. The link is unique to the case and expires in 30 days.
02
Choose your response type
Statement: a brief explanation of the decision process. Documentation: a formal receipt, policy document, or audit log.
03
Submit through the secure page
No account needed. Your response is logged and attached to the public case record immediately.
04
Case is updated
The case status reflects your response. If a receipt was provided, the case closes with the Receipt Available badge.
WHAT A DEFENSIBLE RESPONSE LOOKS LIKE?
A defensible response means your organization can show what happened, where AI was involved, where human review happened, and what proof exists after the decision. Not just a policy — an actual workflow that produces proof.
NOT DEFENSIBLE YET
✕ AI tool makes decision
✕ No log of what was processed
✕ No human review trail
✕ No documentation
✕ Cannot produce proof if challenged
DEFENSIBLE RESPONSE PATH
✓ Decision logged with context
✓ Human review documented
✓ AI role clearly defined
✓ Output is auditable
✓ Can produce proof on request
YOUR QUESTIONS
Does responding mean admitting we did something wrong?
No. Responding demonstrates transparency and a commitment to accountability. It is not an admission of wrongdoing. It is a professional response to a legitimate request.
What if we don't use AI for this type of decision?
You can say exactly that in your response. A clear explanation — even if it is 'AI was not involved' — is valuable and closes the case with a documented record.
Can we dispute the submission?
You can include your full position in the response. If a case contains demonstrably false information, contact hello@audrith.com and we will review it.
What is the EU AI Act exposure here?
High-risk AI systems used for employment, credit, and insurance decisions require documentation under the EU AI Act. Audrith helps you build and demonstrate that documentation. This is not legal advice — consult your counsel.
WHAT WE OFFER
START HERE
$190
Receipt Readiness Review
24–48H ASYNC
Submit your hiring workflow. We identify where proof is missing. No call.
GET STARTED →$1,290
Receipt Gap Audit
5 DAYS
Full mapping of AI-assisted hiring decisions in your process and where receipts are missing.
CONTACT US →$4,890
Fixed Pilot
14 DAYS
We implement your first receipt-ready workflow with full documentation and a response path.
CONTACT US →$2,490/mo
Professional
ONGOING
Ongoing support for decision proofs, response handling, and repeatable receipt workflows.
CONTACT US →ADD-ON — AFTER REVIEW / AUDIT / PILOT
Badge Maintenance
Keep your decision footer, response process, and review cadence active and current each month.
$99–$299/mo