14 sessions.
40 provisions.Two days of learning real-world contracting.

Your customer's data just became training fuel for someone else's model. We unpack the consent, scoping, and usage-restriction language that decides whether the AI provider can learn from your data, share it, or quietly pass it downstream. You'll leave with the exact phrasing to lock the door without killing the deal.

Who owns the output? Who owns the fine-tune? And who gets sued when the model spits out something that looks suspiciously like a competitor's IP? We walk through ownership, license-back, and reuse provisions for AI output and customer-specific customizations, so you can tell exactly what your client is buying.

Modern AI deals route your data through sub-processors, inference partners, and storage layers you've never heard of. We map the architecture the contract is silently authorizing and translate it into the data-flow, sub-processor, and cross-border transfer language your security team will actually approve.

When something goes wrong at 2 am, the contract is the playbook. We compare strong and weak breach-notification, cooperation, and remediation clauses, and rebuild them so notice windows, forensics, and cost allocation are not left to goodwill. Bring your worst breach-clause war story.

Agentic AI takes actions, sends emails, and moves money on your behalf, which means your contract has to scope authority, supervision, and liability for an actor that never reads the contract it is acting under. We draft the guardrails, audit rights, and human-in-the-loop requirements that keep an agent from going rogue on your dime.

The project slips, and suddenly nobody agrees on who was supposed to do what by when. We work through the governance language that keeps a customization project accountable: named project leads, milestone and acceptance mechanics, change-order procedure, dependency and delay allocation, and the escalation path you use before anyone reaches for a termination letter.

Integration work puts each side deep inside the other's systems, and confidentiality drafted for a simple vendor relationship stops holding. We redraw the boundaries: what counts as confidential once it is embedded in a joint build, how residuals and know-how carve-outs actually operate, who the permitted recipients are, and what survives the engagement.

Getting out of a half-finished integration is where weak contracts cost real money. We draft the exit provisions that let your client leave: termination triggers and cure rights, wind-down and transition assistance at a known price, data return and deletion in a usable format, and what happens to work in progress, prepaid fees, and the license to keep running.

When two systems are wired together, the harm rarely traces to one party cleanly. We work through indemnity language built for shared fault: how intellectual property, data, and third-party claims get allocated across a joint build, how contribution and comparative fault are handled, and how defense control, notice, and settlement consent should be written.

A cap, an exclusion, and an exclusive-remedy clause can quietly leave your client with nothing to collect. We take apart the limitation stack: how the cap interacts with carve-outs and indemnities, which damages the exclusion really removes, how super-caps and multipliers are negotiated, and how to keep the remedy you bargained for from being read away.