Evidence by design
Logs, versions and decision records structured to dispute standard now — the difference between proving your case and narrating it.
AI dispute readiness
If AI runs in your business, your next dispute already involves it.
AI dispute readiness from Aun & Co.: evidence, contracts and accountability structured now, so an AI-related conflict finds your business prepared.
Businesses deploying AI acquire a new dispute surface before they notice: vendor promises that may not survive contact with production, outputs that touch customers and third parties, decisions no one can later reconstruct, and logs that rotate away exactly when they become evidence. Dispute readiness is the pre-litigation discipline applied to that surface — building the records, contractual positions and internal accountability now, so that when an AI-related conflict arrives, your business holds the evidence and the high ground rather than searching for both.
The work spans
- Evidence architecture: logs, versions and decisions preserved by design
- Vendor-contract posture: benchmarks, audit rights, liability allocation
- Accountability mapping: who selected, configured and supervised what
- Incident protocol: the first-hours playbook for AI-related harm
- Readiness assessment with a ranked hardening plan
- AI systems make or shape decisions in your business and no log would survive a subpoena.
- Your vendor contracts promise performance in language no one could enforce.
- A customer-facing AI output caused harm and the reconstruction took weeks it should have taken hours.
- The board asked who is accountable for the AI layer and the room went quiet.
The firm works backwards from the disputes AI deployments actually generate — vendor underperformance, harmful outputs, data misuse — and asks of each: could this business prove its case, or its innocence, from the records it keeps today. The gaps become the plan: retention rules that survive system updates, contract amendments that fix benchmarks and audit rights, an accountability map with names on it, and an incident protocol so the first hours of a failure create evidence instead of destroying it.
04 · What you get
Contracts with teeth
Vendor positions hardened while leverage exists: measurable benchmarks, audit rights and liability terms that function under stress.
First hours scripted
An incident protocol that preserves the record and gates the communications — because AI failures are litigated on what happened next.
A typical engagement: a company whose operations lean on several AI vendors commissions a readiness review. The assessment finds unprovable performance promises and logs that rotate monthly; within a quarter, contracts are amended, retention is fixed, and an incident protocol exists — before any dispute tests them.
Described in abbreviated, anonymised form to preserve client confidentiality.
What is AI dispute readiness?
The pre-litigation hardening of an AI-deploying business: evidence retention designed for future proof, vendor contracts with enforceable benchmarks and audit rights, a named accountability map, and an incident protocol. It is posture work — done before the dispute chooses the timing for you.
What evidence should businesses keep about their AI systems?
Enough to reconstruct any consequential decision: inputs and outputs at issue, model and configuration versions, override and supervision records, and the vendor's representations over time. Retention must outlive system updates — the update cycle is where AI evidence quietly dies.
How does a readiness assessment work?
The firm reviews your AI deployments, vendor contracts and record-keeping against the disputes they realistically generate, and returns a ranked hardening plan — evidence, contracts, accountability, protocol. The assessment runs in weeks; the hardening is sequenced by exposure.