When it comes to enterprise AI, experimentation is out — and operationalization is in. KPMG’s latest AI Quarterly Pulse Survey of more than 130 C-suite leaders at billion-dollar companies shows that full deployments of AI agents have nearly tripled, jumping from 11% in Q1 to 33% in Q2 2025.
It’s the clearest signal we have seen among all of the AI hype that organizations are actually moving well beyond pilots and proofs-of-concept, embedding AI agents as digital employees in core business workflows.
🚀 From Pilot Projects to Production Scale
KPMG’s data shows this shift is happening fast:
- The share of firms still in pilot phases has fallen to 57%, down from 65% just three months ago.
- Companies that are only exploring or researching dropped from 25% to 10%, a steep decline that highlights a growing sense of urgency.
- On average, firms plan to invest $114 million in GenAI initiatives over the next year — up 28% from the $89 million projected in Q1.
⚖️ Efficiency and Growth — Not Just Cost Cutting
With much of our early focus at NewTide on efficiency and automation of routine tasks it was eye opening to see how leaders view of the why behind AI agents is rapidly evolving from just saving money to making money. From a focus on efficiency to a focus on revenue growth:
- 46% say they’re deploying AI agents to drive both operational efficiency and new revenue opportunities.
- 93% of executives agree that their early GenAI investments have already improved their competitive positioning.
This balance — pairing automation and cost savings with top-line growth — echoes what we are starting to see across the NewTide client community: as AI agents succeed at back-office optimization; they unlock new value streams by working alongside people in day-to-day operations who are freed up to pursue growth.
🔒 Trust, Oversight & Human-in-the-Loop Are Non-Negotiable
But moving to full deployment doesn’t mean full autonomy. According to the survey:
- 55% of deployed agents still require human-in-the-loop (HITL) oversight.
- 45% of firms won’t allow agents to access sensitive data without human review.
- Top concerns include cybersecurity risks (38%), misinformation, accuracy, and data quality.
This governance-first approach underlines what we believe at NewTide.ai: trust is the foundation for scaling AI. Our frameworks and workflows keep humans accountable, data protected, and decisions explainable — so you can scale AI responsibly, not recklessly.
⚙️ Embedded AI Is Becoming Routine
Daily usage is rising across multiple enterprise AI tools:
- Use of AI productivity tools surged from 22% to 58% in just one quarter.
- Weekly usage of knowledge assistants climbed from 48% to 61%.
- Embedded GenAI workflows are now active in 35% of organizations, up from 24%.
The implication: AI agents are no longer an experiment on the sidelines — they’re quickly becoming integral digital teammates.
🧭 What This Means for You
The story behind these numbers is clear: the AI adoption curve has bent sharply upward — and leaders are aligning budgets, governance, and human oversight to make AI agents work at scale.
At NewTide.ai, we help organizations move beyond pilots by deploying AI agents that:
✅ Tap directly into your existing data and systems
✅ Operate with transparent guardrails and audit trails
✅ Support your teams with explainable recommendations — never black boxes
As the KPMG survey shows, the next wave of value comes from pairing robust automation with trust and control.
🔗 Further Reading & Resources
📄 Read the full KPMG Q2 2025 AI Pulse Survey
📄 Dive into the KPMG Press Release
Want to explore how your organization can scale AI responsibly? Let’s talk about how NewTide’s Agent Harbor, Data Helm, and Shipyard platform helps you deploy digital employees with the same trust, oversight, and governance you expect from your human teams.
