Layer researchExperience and channels
Experience & channels
Status: complete (published 2026-08-19). Scope finalized at Batch D kickoff with maintainer POV; see brief.md, findings.md, vendors.md, sources.md.
Scope inventory (v0)
- Estate: web and mobile assistants, contact center (voice, agent-assist, deflection), commerce, marketing automation
- Trust and liability: groundedness bars, AI disclosure rules, action authority limits (refunds, commitments), hallucination liability precedents, consent-based customer memory
- Hardening: the public prompt-injection surface, fraud and abuse, rate limiting, authenticated vs anonymous capability tiers
- Machine customers: inbound bot-to-bot traffic, agent-readable storefronts and APIs, emerging agentic payment protocols
Challenged-default candidates
Full conversational front doors vs scoped task agents; CCaaS vendor AI vs bring-your-own agents. Each track proposes its final list at kickoff.
Files
brief.md, findings.md, vendors.md, sources.md are created from ../_TEMPLATE/ at kickoff.
Source: research/R09-experience-and-channels/README.md in the evidence repository behind this site.