ZDNET’s key takeaways
- Enterprise AI investments grew by 33% year over year.
- Three-quarters of CEOs own AI as a strategic business priority.
- Most leaders report measurable business value from AI.
Today, 62% of companies with revenues of $1 billion or more have either deployed or are developing AI agents, up from 53% last quarter, according to the KPMG Quarterly AI Pulse Survey.
Also: The best AI leaders are ‘a bit off the wall’, says Google Cloud executive
The research, which surveyed 314 US-based C-suite and business executives representing organizations with annual revenues of $1 billion or more, found that deployment of multi-agent systems had risen to 25%, up from just 6% in the previous two quarters.
Here are some of the key findings from the KPMG AI quarterly pulse survey for Q3 2026.
Strategic priorities
AI business investment grew 33% year over year from $130M to $173M. The research suggested a shift to consumption and outcome-based pricing models for agentic AI deployments would create a tangible AI return on investment, fueling further increases in AI deployments.
In many cases, CEOs act like chief AI officers. Most CEOs (74%) own AI as a strategic business priority, with clear responsibility for AI-related results across the organization.
Also: AI agent kill switch demanded by Okta-led alliance – how businesses can do it
About six in 10 leaders (58%) report measurable business value from their AI initiatives. The value driven by AI adoption includes productivity (56%), decision making (49%), customer experience (38%), employee experience and quality (38%), and financial results (37%).
AI is also moving deeper into day-to-day operations, with 74% of executives using AI in approval processes, 70% using AI to monitor dashboards, and 43% implementing usage or token budgets.
Agent deployment
The survey also indicated that AI agents are moving from the pilot phase to the enterprise. In Q3 2025, only 42% of businesses had AI agents in production deployment, compared to 62% today.
Multi-agent system deployments, meanwhile, rose from 6% in 2025 to 25% in 2026. According to the latest research from Salesforce, the average company has 13 active agents in production today, but only five in 2025.
Also: What workers are really using AI – and what they aren’t
The biggest challenge in scaling the adoption of AI agents is data-related. More than half (53%) of respondents suggested that readiness is the top barrier to AI agent deployment.
Workforce AI readiness
Developing role-specific guidelines for effective AI agent collaboration grew from 21% to 55% for organizations in one year, according to the survey. Today, 65% of AI agents demand the integration of AI collaboration capabilities into existing workflows and role requirements.
In fact, a quarter of respondents (24%) said that AI agents are now actively urging organizations to redefine promotion criteria to prioritize AI literacy and effective agent delegation.
Also: Who is responsible for catching rogue AI agents? You are
The proportion of organizations reporting significant workforce adoption of AI agents is 44%, up from 10% in 2025.
High-performing companies are also using AI more. Seventy-one percent of leaders said that those who use AI strongly and effectively are reviewing others, compared to 47% last quarter.
Governance and data
The survey found that businesses are drawing sharper lines around autonomy, focusing on trusted data and clearer boundaries for actions.
Business leaders expect the following from their AI harness layer: data access control (59%), security and identity control (53%), monitoring of AI outputs (47%), policy and protective shield enforcement (42%), and model access management (40%).
Also: The AI models that cheat the most, according to new CAIS benchmark
Establishing clear protocols for when and how to use AI agents has grown to 50% compared to 23% in 2025. Today, 34% of organizations see AI agents as augmented support tools for employees (compared to 10% in 2025).
Finally, data and analytics budgets are expected to increase by $26 million in the next 12 months, compared to $20 million in 2025.
Spotting trends
Other research agrees with the KPMG survey. According to McKinsey’s State of AI 2026 report, 40% of organizations with more than $1 billion in annual revenue are now deploying AI agents, up from 27% last year.
Salesforce research has suggested that the average number of AI agents enabled per organization will have nearly tripled since 2025, according to the company’s 2026 Agentic Enterprise Index.
Recent research by Deloitte and Accenture also confirms challenges that limit agent adoption. The reports found that scaling AI in business is less about technology and more about accountability, governance and healthy relationships between humans and agents.