Agentforce World Tour London 2026<\/a> to push a single argument: the first wave of AI agents drove efficiency, and the second wave will drive revenue. The event showcased Thames Valley Police’s “Bobbi” agent handling 14,000 citizen contacts annually while freeing 3,266 officer hours, Capita deploying Agentforce across defence and education under a new Framework Agreement, and the Fin acquisition at roughly $3.6 billion closing later this year. Slackbot’s MCP client, which connects 20-plus partner apps into one conversational interface, is now generally available. Salesforce’s own survey found 61% of UK workers have received no formal AI training from their employers.<\/p>\nWhat this means for your business<\/h2>\n
The Fin acquisition is the signal CIOs should be watching most carefully. Salesforce is buying a purpose-built customer support AI model, Apex, that reportedly outperforms GPT and Claude on resolution rates, which means Salesforce is no longer content to orchestrate other vendors’ models at the customer-facing layer. If your AI strategy currently relies on Agentforce as a workflow coordinator sitting on top of commodity LLMs, that architecture is about to get more opinionated. Organisations already deep in the Salesforce stack gain a tighter vertical integration; those who chose Salesforce for CRM and a different vendor for AI agents will face renewed pressure to consolidate.<\/p>\n
The Headless 360 announcement deserves a second look beyond the marketing framing. The architecture separates the reasoning layer from the rendering layer, meaning agents can surface outputs natively in Slack, WhatsApp, or voice without custom integration work for each channel. The MCP standard underneath this, a protocol letting AI models request data and actions from external tools the way a browser requests web pages, is genuinely open. That matters because it gives Salesforce a credible answer to the “we don’t want to be locked in” objection while still making its own platform the default orchestration hub. The lock-in moves one layer deeper, from the application to the data context, where Salesforce already holds the advantage through its CRM records.<\/p>\n
The 61% figure on AI training is the number that should force a budget conversation CIOs have been deferring. Salesforce has an obvious commercial interest in framing this as a readiness gap rather than an adoption gap, and its survey of UK workers was commissioned to support that narrative. But the underlying dynamic is real: most enterprise AI rollouts are stalling not at the model layer but at the human layer, where staff either distrust the output or don’t know how to act on it. The organisations that close this cycle fastest, giving workers enough training to act confidently on agent recommendations, are the ones that convert efficiency gains into the revenue-generating use cases Salesforce is now selling. That gap is where this story is actually about you.<\/p>\n
Concept deep-dive: Model Context Protocol (MCP)<\/h2>\n
MCP is an open standard that lets an AI model ask external tools for data or actions during a conversation, the way a spreadsheet formula pulls live data from a connected source rather than relying on whatever was pasted in. Instead of custom connectors built separately for each app, one standard interface handles all of them. For enterprise buyers, the business consequence is that the platform controlling MCP routing, in this case Slack acting as the orchestration layer, becomes the chokepoint through which every agent interaction travels.<\/p>\n
Based on reporting from Agentforce World Tour London ’26 Live: All the Highlights You Missed<\/a>, originally published 2026-06-18 03:00:00.<\/em><\/p>\n","protected":false},"excerpt":{"rendered":"Share with your CIO Salesforce used its Agentforce World Tour London 2026 to push a single argument: the first wave of AI agents drove efficiency, and the second wave will drive revenue. The event showcased Thames Valley Police’s “Bobbi” agent handling 14,000 citizen contacts annually while freeing 3,266 officer hours, Capita deploying Agentforce across defence […]<\/p>\n","protected":false},"author":1,"featured_media":5399,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[148],"tags":[185],"tmauthors":[],"class_list":["post-5398","post","type-post","status-publish","format-standard","has-post-thumbnail","category-ai-marketing","tag-cio"],"_links":{"self":[{"href":"https:\/\/workai.tv\/news\/wp-json\/wp\/v2\/posts\/5398","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/workai.tv\/news\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/workai.tv\/news\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/workai.tv\/news\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/workai.tv\/news\/wp-json\/wp\/v2\/comments?post=5398"}],"version-history":[{"count":0,"href":"https:\/\/workai.tv\/news\/wp-json\/wp\/v2\/posts\/5398\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/workai.tv\/news\/wp-json\/wp\/v2\/media\/5399"}],"wp:attachment":[{"href":"https:\/\/workai.tv\/news\/wp-json\/wp\/v2\/media?parent=5398"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/workai.tv\/news\/wp-json\/wp\/v2\/categories?post=5398"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/workai.tv\/news\/wp-json\/wp\/v2\/tags?post=5398"},{"taxonomy":"tmauthors","embeddable":true,"href":"https:\/\/workai.tv\/news\/wp-json\/wp\/v2\/tmauthors?post=5398"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}