AI Agent News Today
Tuesday, June 23, 2026DaVinci Commerce launches Agentic BrandStore Enterprise
What changed: DaVinci Commerce announced general availability of Agentic BrandStore Enterprise — a no-code platform that enriches product catalogs with conversational metadata and can deploy branded storefronts optimized for AI-agent discovery and purchases across major LLM platforms.
Why it matters: For brands and merchants this is a turnkey way to make product content machine-readable for AI agents (the places consumers now ask buying questions), reducing the manual rework of catalogs and speeding time-to-discoverability on chat platforms. That can materially improve how often AI-driven queries surface your products and whether those sessions convert to sales.
Try/watch: If you sell online, audit one product category for conversational discoverability this week: check how the top 10 buyer questions map to your SKU copy, then pilot an enrichment feed or third‑party BrandStore connector and measure discovery traffic lift and conversion. Watch for vendor claims about “instant” integrations — validate which AI platforms and review/data sources they actually support.
Moneris coverage: Canadian payments firm positions an MCP server for agentic commerce
What changed: Coverage reports that Moneris launched a Model Context Protocol (MCP) server to let AI agents securely interact with Moneris payment APIs — a connector designed to let developers integrate once and expose payment capabilities safely to MCP‑compatible AI ecosystems.
Why it matters: If you run commerce in Canada (or integrate with Canadian payments), an MCP server from a major processor reduces custom engineering to support agentic checkouts and keeps payment controls and fraud protections inside the merchant’s existing stack — an important trust and compliance consideration when agents initiate transactions.
Try/watch: Product and payments teams should review Moneris’ MCP developer docs and permission model before enabling any agentic checkout flows. Run tabletop threat and fraud scenarios (refunds, credential misuse, supply‑chain data leakage) and require explicit approval controls and audit logs from any MCP connector you use. Watch how processors publish their MCP permission scopes and attestation guarantees.
Do not just read about agents. Build one that runs.
Create an agent from a short prompt, connect a gateway later, and pay mainly for active runtime.
Hosted agent
OpenClaw or Hermes