Overview

• Large-company AI adoption has edged down from a peak of 14% to 12%, per the U.S. Census BTOS, after an MIT survey reported about 95% of generative-AI pilots at big firms were failing.
• Vendors are accelerating launches: HubSpot rolled out 15 agents and a marketplace, Adobe announced six agents with an agent composer, Airtop introduced a conversational agent builder, Unily released an Agent Orchestrator, and Grove AI debuted a clinical-trials agent.
• Salesforce CEO Marc Benioff said he cut 4,000 customer service roles as AI agents reduced support time and costs, underscoring near-term labor impacts even as many companies prioritize human oversight.
• Analysts and executives stress guardrails, auditing, and human-in-the-loop controls, with HBR and Gartner highlighting organizational readiness and a forecast that about one-third of enterprise apps will include agentic AI by 2028.
• Healthcare remains a leading testbed, with pilots in claims adjudication, prior authorization, and care coordination informed by Stanford–CMU research showing agentic systems outperform standard agents by 48% on complex reasoning tasks.