By Patrick Miller
October 12, 2026

The newest worker in your company needs no visa, no desk, and no sleep. It’s the AI/SI agent, and the numbers behind it are already big. Gartner predicts agentic AI/SI will resolve 80% of common customer-service issues without human intervention by 2029, cutting operating costs 30%. [1] Vendors already price the work by outcome: $0.99 per resolution at Intercom, and $2 per conversation under Salesforce’s conversational pricing. [2][3] Gartner also expects 33% of enterprise software applications to include agentic AI/SI by 2028, up from less than 1% in 2024, and at least 15% of day-to-day work decisions to be made through it, up from 0%. It says that by 2027, over 65% of engineering teams using agentic coding will treat traditional development environments as optional. [4][5]

Now the honest counterweight. Gartner predicts over 40% of agentic AI/SI projects will be canceled by the end of 2027, due to escalating costs, unclear business value, or inadequate risk controls. [4] It estimates only about 130 of the thousands of vendors selling “agentic” products are real, and calls the rest “agent washing.” [4] Its advice is to pursue agents only where they deliver clear value. The pattern is clear enough to act on. Agents do best on high-volume, rules-bound work with a clear finish line: answering common questions, updating records, routing requests, drafting first versions. They struggle with ambiguous goals, nuanced instructions held over time, and work where a wrong answer is expensive. [4]

For operators, that points to a simple playbook. Start narrow, with one workflow where success is easy to count. Measure outcomes, not activity: resolutions, error rates, cost per finished task. Grow an agent’s permissions only as its reliability is proven. And keep a person approving anything that commits money, signs a contract, or speaks publicly for the company. The shift also creates new roles: people who brief agents, review their work, handle the exceptions, and own the results. The same agents are starting to direct robots, drones, and machines, and the rule carries over there too: a person approves.

This part is my opinion, not fact. Costs are likely to fall further than outsourcing ever took them: outsourcing cut the price of an hour; agents cut the number of hours. Value is likely to shift toward the technology companies that own the models, clouds, and compute, rather than toward firms that sell hours of labor. If agents are the new tireless workers, people’s value is judgment and accountability: deciding what good looks like, and answering for it when it’s wrong. A machine can do the work, but it can’t be held to account for it. That’s why the most important button in this new workplace says approve, and a person should always be the one to press it.

Company names are used for identification only. Logical Solutions Group is not affiliated with or endorsed by any company named here.

Sources

  1. Gartner, “Gartner Predicts Agentic AI Will Autonomously Resolve 80% of Common Customer Service Issues Without Human Intervention by 2029,” Mar 5, 2025. https://www.gartner.com/en/newsroom/press-releases/2025-03-05-gartner-predicts-agentic-ai-will-autonomously-resolve-80-percent-of-common-customer-service-issues-without-human-intervention-by-20290

  2. Fin (Intercom), “Fin pricing: Outcomes,” help center, accessed Oct 9, 2026. https://fin.ai/help/en/articles/13975800-fin-pricing-outcomes

  3. Salesforce, “Salesforce Introduces New Flexible Agentforce Pricing,” May 15, 2025 (”existing conversational pricing at $2 per conversation”). https://www.salesforce.com/news/press-releases/2025/05/15/agentforce-flexible-pricing-news/

  4. Gartner, “Gartner Predicts Over 40% of Agentic AI Projects Will Be Canceled by End of 2027,” Jun 25, 2025 (also the source for the 33%-of-applications and 15%-of-decisions forecasts, the ~130 real vendors estimate, and the limits on complex goals and nuanced instructions). https://www.gartner.com/en/newsroom/press-releases/2025-06-25-gartner-predicts-over-40-percent-of-agentic-ai-projects-will-be-canceled-by-end-of-2027

  5. Gartner, “Gartner Says the Market for Enterprise AI Coding Agents Is Entering a New Phase of Expansion and Competitive Realignment,” May 20, 2026. https://www.gartner.com/en/newsroom/press-releases/2026-05-20-gartner-says-the-market-for-enterprise-ai-coding-agents-is-entering-a-new-phase-of-expansion-and-competitive-realignment