A confidential benchmarking snapshot for our Annual Meeting in Seattle | July 30 Agentic AI Survey (Anonymized Chatham House summary results available July 30, 25)Name*1. Where is AI in your Enterprise IT agenda?*Important but not top-tierTop 10 priorityTop 3 priority2. Where are you seeing impact from AI - Today?*(Select as many as you like)Individual productivity (e.g. developer or analyst copilots)Team-level workflows (e.g. agents coordinating tasks across users)Department-level automation (e.g. finance supply chain service ops)Enterprise-wide transformation3. Where are you in the journey with AI Agents or Agentic AI?*Planning / Getting StartedRunning live pilots or MVPsDeployed in productionScaling AI agents or Agentic AI4. What percent (0-100) of your total AI efforts (predictive + generative + agentic = 100) is Agentic AI ?*Note: (1) “Agentic AI” refers to systems that autonomously plan, reason, and act across multi-step workflows toward a goal. This is distinct from assistive-only AI agents like chatbots or copilots. (2) Effort is specific to where you are investing your mindshare today (not necessarily in production)5. How do you view the potential impact of agentic AI over the next 3 years?*Game-changing — foundational to how we operateHigh potential — but requires major redesignUseful in targeted domains onlyOverhyped — not materially transformativeToo early to say6. Which functions or processes are top targets for agentic AI?*(and proactively targeted - as opposed to LLMs eventually becoming part of every application in the enterprise in some form)(Select as many as you like)Software engineering / DevOpsCustomer service and supportSales and marketingFinance and procurementCybersecuritySupply chain and logisticsHR and internal knowledge managementIT operations and observabilityOther7. How are you currently deploying or planning to deploy agentic AI?*(Select as many as you like)Embedding basic agents in SaaS platforms (e.g. Salesforce ServiceNow)Extending RPA or automation tools with LLM-based decisioningBuilding standalone agents using LLM APIsAssembling multi-step agents using orchestration frameworks (e.g. LangChain Semantic Kernel)Coordinating multiple agents across workflows (e.g. agent-to-agent systems Model Coordination Platforms)Not yet deploying — still exploring8. Who are your primary foundation model providers or partners?*Please pick the models you use independent of the platform you run them on, eg Microsoft Fabric AI / AWS Bedrock)(Select as many as you like)OpenAIGoogle / GeminiAnthropic / ClaudeMeta / LlamaMistralHugging Face / other open source9. What is largest barrier to scaling agentic AI in your organization?*(Select as many as you like)Organizational readiness (skills structure mindset)Finalizing corporate strategy on AgencyIntegration with legacy systems and data silosGovernance and accountabilityCost or unclear ROIStill defining use cases and value10. How would you describe your organization’s readiness to scale agentic AI?*Too early to assessNot ready — significant redesign requiredPartially ready — building technical and organizational capabilitiesReady — strategy architecture and operating model in placeSubmit
