Only 1 in 5 companies has a mature model for overseeing autonomous AI agents. That is one of the findings in Deloitte’s State of AI in the Enterprise 2026 report, based on 3,235 leaders across 24 countries. AI adoption is accelerating, but governance is not keeping pace. Here is what that gap means for the people running operations and fleets.
From pilots to real production
AI is moving past the isolated-experiment stage. Worker access to AI tools grew 50% during 2025, and the share of companies with at least 40% of their projects in production is set to double within six months. The outcomes organizations report back this up:
- 66% see gains in productivity and efficiency.
- 53% make better decisions from richer data.
- 40% cut costs and 20% already see revenue growth.
The bottleneck is governance, not technology
The same study shows the flip side: only 20% of organizations have a mature oversight model for agents that act on their own. The skills gap stands out as the biggest barrier to integration, and just 34% are using AI to fundamentally rethink their business, while the rest stay at surface-level improvements. Scaling without clear rules invites errors, downtime, and decisions that are hard to audit.
What it means for fleet management
For a fleet manager, the promise of agentic AI is tangible: systems that prioritize alerts, anticipate maintenance, and suggest route adjustments without constant supervision. But no agent decides well on incomplete data. The starting point is clean, reliable, real-time information from every vehicle, paired with explicit rules about what the system may decide and what stays with a human.
The EsencialGPS platform brings together real-time location, sensor data, and alerts from every vehicle, the ordered foundation any AI layer needs to make sound decisions. Learn more at www.esencialgps.cl.
Based on data from Deloitte.
