In this live Q&A, TSIA's George Humphrey walks through TSIA's latest research on agentic AI in support, and Inbenta CTO Merlin Bise answers from the vendor seat: where assembled stacks break first, what a specialist should own, and what ready actually looks like.

‍

Why you should attend:

‍

• Research, not a pitch. The discussion draws on TSIA's 2026 Agentic AI in Support Services survey and its build, buy, or assemble framework.
• The ownership question, answered directly. Who governs, measures, and pays for the AI your customers talk to.
• A candid view from the vendor side. Merlin on the failure patterns he sees most often.

‍

You'll leave with:

‍

• A five-layer operating model that puts a named owner on standards, the agent, the seam, quality, and economics.
• Five moves to make before adding another tool, starting with one that costs nothing: naming the owner of every production agent.
• The metrics that grade an assembled stack, including AI-isolated CSAT, autonomous resolution rate, and actual cost per resolution.

‍

Speakers:

Merlin Bise leads technology at Inbenta and has long argued that a platform trying to do everything delivers average AI. His test is simple: if your best engineers are building search, chat, or voice and that isn't your core business, they're building the wrong thing. He is joined by George Humphrey, Sr. VP of Research at TSIA.

AI adoption is moving faster than the contracts behind it. Liability gaps in vendor agreements, privacy obligations under GDPR and CCPA, and unsettled questions about who owns AI-generated content are landing on CX leaders' desks, often without in-house counsel in the room. Inbenta Chief Legal Officer Adam Rivera translates that risk into plain-language frameworks you can use before you sign. No legal jargon, no sales pitch.

‍

Why you should attend:

‍

• The rules are still moving. The EU AI Act, new US state laws, and shifting case law on IP and liability all shape decisions you're making now.
• Built for buyers without a legal  team on call. Designed for Director-level and above leaders in CX, customer success, and technology services who evaluate, buy, or govern AI.
• Grounded in real procurement work. Examples come from actual AI contracting and compliance programs, not hypotheticals.

‍

You'll leave with:

‍

• A plain-language risk framework covering AI liability, data privacy, IP ownership, and vendor accountability, and how each affects your real exposure.
• The contract terms to treat as non-negotiable, including indemnification provisions and audit rights.
• A repeatable due diligence checklist, with the questions to ask any AI vendor before you sign.

‍

Speaker:

Adam Rivera has spent his career building the privacy and compliance programs that sit behind large-scale data and technology businesses. Before joining Inbenta as Chief Legal Officer and Corporate Secretary, he led privacy and compliance at DoorDash, Refinitiv (LSEG), and Thomson Reuters. His work spans privacy law, data governance, cybersecurity, and cross-border M&A, and he holds a JD and an MBA.