Gartner Symposium 2026: AI Agents, Risk and Value Data

Short answer: Gartner opened its IT Symposium/Xpo in Orlando on October 19, 2026, with a blunt message: AI value is growing, but AI risk is growing faster. Only 13% of CIOs said their enterprise has realized significant or transformational value from AI, and Gartner predicted that by 2028, 25% of enterprise security breaches will trace back to AI agent abuse (Gartner Day 1, Gartner Day 2).

Key Takeaways

13% of CIOs report significant or transformational AI value.
11% of CIOs qualify as "AI Vanguards" in the 2027 Gartner CIO and Technology Executive Survey.
25% of enterprise breaches will trace back to AI agent abuse by 2028, Gartner predicts.
44% of CEOs say it is difficult to show AI-driven financial returns to stakeholders.
$7.35 trillion: forecast worldwide IT spending in 2027, up 15%.

CIOs reporting significant or transformational AI value: Figure: 13%; Source: Gartner, October 2026
CIOs who are AI Vanguards: Figure: 11%; Source: 2027 Gartner CIO and Technology Executive Survey
CEOs who find it difficult to show AI financial returns: Figure: 44%; Source: Gartner CEO Watch 2026
CEOs planning to increase technology investment: Figure: 75%; Source: Gartner CEO Watch 2026
CEOs naming operating model redesign as a top move for AI returns: Figure: 63%; Source: Gartner CEO Watch 2026
Worldwide IT spending, 2027: Figure: $7.35 trillion, up 15%; Source: Gartner forecast

What Did Gartner Say About AI Value and Risk?

In the opening keynote, Gartner analysts Kristin Moyer and Daryl Plummer said that enterprises' need to build AI value is strong, but their ability to govern the risk is weak. Moyer proposed an "AI central bank" to support AI governance and accountability. On Day 2, Mary Mesaglio argued that the AI value gap is really an AI risk gap and that cost is the biggest tactical risk to scaling AI.

Gartner's AI Vanguards, the 11% of CIOs getting leading results, are more likely to set explicit criteria for stopping pilots and abandoning low-value initiatives. Gartner said 3 gaps hold other CIOs back: cost management, value and IT workforce.

What Did Gartner Say About AI Agents?

Gene Alvarez predicted that 25% of enterprise security breaches will trace back to AI agent abuse by 2028, citing unmanaged machine privileges and legacy credential practices. He said AI agent identity controls can replace high-risk static credentials with dynamic, policy-controlled access. Svetlana Sicular advised building a layered guardrails stack, with oversight matched to each agent's level of autonomy, and collecting evidence now to prepare for external scrutiny.

Mesaglio said AI agents are "experiencing a reality check," with the workable zone sitting where reliability and security meet autonomy. Gartner also named the Autonomous Agent Platform as 1 of 3 themes in its 2027 strategic technology trends, alongside the Non-Textual AI Frontier and Durable Agility.

Enterprise breaches traced to AI agent abuse: Timeline: 25% by 2028; Analyst or source: Gene Alvarez, Gartner
Enterprises abandoning agentic AI built by vendor FDE: Timeline: 70% by 2028; Analyst or source: Mukul Saha, Gartner
Enterprises that will be AI-first: Timeline: More than 1 in 10 by 2030; Analyst or source: Gartner
Physical AI makes other forms of AI legacy: Timeline: By 2035; Analyst or source: Mary Mesaglio, Gartner

How Does This Connect to Gartner's Warning on Forward-Deployed Engineering?

3 weeks before the Symposium, Gartner predicted that by 2028, 70% of enterprises will abandon agentic AI built through vendor forward-deployed engineering because of rising costs and an inability to evolve the systems independently. Analyst Mukul Saha warned of "FDE washing," where consulting services are relabeled as forward-deployed engineering, and recommended contract terms covering knowledge transfer, IP rights and exit plans.

What Should Enterprises Do Next?

Set stop criteria for every AI pilot. Vanguards define in advance when a pilot ends.
Give every agent an identity. Replace shared and static credentials with scoped, dynamic access.
Match oversight to autonomy. Read-only agents need lighter controls than agents that change systems or move money.
Track cost per task. Gartner calls cost the biggest tactical risk to scaling AI.
Plan the exit on day 1. Any external build should end with your team able to run and change the system.

Frequently Asked Questions

What did Gartner say at IT Symposium 2026?

Gartner's opening keynote argued that AI value is growing but AI risk is growing faster. Only 13% of CIOs said their enterprise has realized significant or transformational value from AI.

What did Gartner predict about AI agents?

Gartner predicted that by 2028, 25% of enterprise security breaches will trace back to AI agent abuse, driven by unmanaged machine privileges and legacy credential practices.

What are Gartner's 2027 strategic technology trends?

Gartner grouped its 2027 trends into 3 themes: the Non-Textual AI Frontier, the Autonomous Agent Platform and Durable Agility.

What is an AI Vanguard CIO?

Gartner's term for CIOs whose organizations get leading results from AI. Only 11% of CIOs qualify, according to the 2027 Gartner CIO and Technology Executive Survey.

Planning an AI agent or consulting project? Book a consultation with Deployed Labs.