• The AI Underground With CAIS
  • Posts
  • H1 2025 Insights, H2 2025 Imperatives, and 2026 Predictions: How Agentic AI Will Define the Next Competitive Edge

H1 2025 Insights, H2 2025 Imperatives, and 2026 Predictions: How Agentic AI Will Define the Next Competitive Edge

If you're wondering where the next competitive edge lies or how your business will operate in H2 2025 and in 2026, read on. We answer questions such as: Did a country really give AI citizenship? What did H1 2025 reveal about the real-world impact, and ROI, of AI agents/the agenticOS across industries? As we move into H2, are you prepared for a future where operations coordinate themselves and AI agents function as teammates, or even leaders? And as 2026 approaches and we get into 2026, are you ready? Are you ready to face the strategic, legal, and ethical questions surrounding AI identity, autonomy, and accountability? Read on for this and much more...

Author

Ross Green
July 15, 2025

Table of Contents

Introduction: The Inflection Point Is Here — Welcome to the Era of the agenticOS/Agentic AI

H1 2025 has taught us that we are currently living in the age of agentic AI/the agenticOS, no longer simply peering into it from afar.

To (re)introduce the idea, CAIS’ agenticOS is a novel swarm of AI agents (i.e., sales, marketing, project management, HR, etc) attached to a one-source-of-truth agenticDatabase designed to break down the artificial barriers of various departments to create interdepartmental employees. Check out more information here.

In just the first half of 2025, nearly 95% of U.S. companies have adopted generative AI, and for many, the results are transformative. We can see the ROI now having gone through the first half of 2025 (check out this article for a short synopsis of H1 2025). Read on in this newsletter for much more data to demonstrate the current state of agentic AI post-H1 2025.

But what’s most exciting isn’t just the scale of adoption; rather, it’s really the shift in how AI is being used. We’re witnessing the dawn of autonomous agentic AI: intelligent systems that act, learn, remember, and collaborate.

This newsletter unpacks the defining trends of H1 2025 by the numbers and the accelerating forces shaping H2 and beyond. From multimodal AI agents that can see, hear, and speak, to the emergence of continuous agentic workflows that reduce handoffs and micromanagement, to the rise of a full-fledged agent marketplace reminiscent of the early App Store era, only with far greater implications, the agenticOS and agentic AI are becoming foundational as we go from H2 2025 to 2026.

As such, as 2026 gets underway, we predict that deeper and more urgent questions are gaining momentum: What rights, if any, should AI agents have? Should they possess persistent identities? And are we approaching a future where some organizations might be led by AI agents, not merely supported by them? (Spoiler: an AI CEO exists already).

So if you're wondering where the next competitive edge lies or how your business will operate in H2 2025 and in 2026, read on.

Type of Metric

H1 2025 Findings

Reference

Generative AI Adoption (US)

95% of companies in the US are using generative AI, a 12 percentage-point increase in one year.

1

ROI Realization

Nearly 74% of enterprises report that their most advanced generative AI initiatives are meeting or exceeding ROI expectations.

2

Adoption Focus (Action vs. Chat)

Over 70% of enterprise AI efforts in 2025 center on action-based AI agents (automation that takes action), rather than just conversational chatbots.

3

Customer Service Cost Savings

AI virtual agents can cut customer service costs by up to 30%.

4

Customer Service Productivity

Organizations using generative AI agents in support saw a 14% increase in issues resolved per hour and a 9% decrease in handling time.

5

Sales Process Impact

About 68% of sales professionals say AI-powered predictive analytics have transformed how they identify high-potential prospects.

6

Marketing Adoption

88% of marketing professionals are now using AI in their day-to-day roles.

7

HR Adoption

Approximately 65% of HR departments globally have implemented at least one AI-based tool as of 2025.

8

Enterprise AI Agents Value

79% of surveyed enterprises have adopted AI agents, and among those adopters 66% report measurable productivity gains from these agents. (Additionally, 88% of executives plan to boost AI budgets in the next year due to agentic AI.)

9

Fortune 500 Adoption

90% of Fortune 500 companies have already used Microsoft’s Copilot Studio to build AI agents and automations.

10

References: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10

Explained: In the first half of 2025, the adoption and impact of generative AI across U.S. enterprises has reached an inflection point. A staggering 95% of companies in the U.S. are now using generative AI, marking a 12-point increase from the previous year and signaling that AI has moved from experimentation to near-universal implementation. Among those using it, the payoff is becoming clear: 74% of enterprises report that their most advanced AI initiatives are meeting or exceeding ROI expectations, demonstrating that these technologies are not just hype but are delivering tangible business value.

Notably, the focus of AI adoption has shifted significantly. While early efforts often centered around conversational tools like chatbots, over 70% of enterprise AI deployments in H1 2025 are centered on action-based agents, or systems that are proactive rather than reactive. These include tools that automate workflows, initiate tasks, and make real-time decisions, marking a major evolution from passive interaction to proactive execution. In customer service, for example, AI virtual agents are reducing support costs by up to 30%, and those leveraging generative AI in their support operations have seen a 14% increase in issues resolved per hour along with a 9% decrease in average handling time; these are clear indicators of improved efficiency and scalability.

The impact extends well beyond support functions. In sales, 68% of professionals say that AI-powered predictive analytics have transformed how they identify high-potential prospects, shifting sales from reactive to data-driven and anticipatory. Marketing teams are also deeply immersed in AI, with 88% of marketing professionals now using AI tools in their daily work, embedding intelligence across campaign planning, content creation, and customer targeting. Similarly, HR departments worldwide have joined the wave, with 65% now using at least one AI-based tool, helping streamline everything from recruiting to employee engagement.

Perhaps most telling is how enterprises are institutionalizing these capabilities. 79% of surveyed companies have adopted AI agents, and two-thirds of them are already reporting measurable productivity gains. This success is driving further investment, with 88% of executives planning to increase their AI budgets over the next year, largely in response to the promise of agentic AI. This trend is echoed at the highest levels of enterprise: 90% of Fortune 500 companies have already used Microsoft’s Copilot Studio to build custom AI agents and automations, highlighting how mainstream these tools have become. Altogether, these findings paint a picture of AI not as a future possibility, but as a present-day engine of transformation across nearly every business function.

Type of Metric

Brief Findings

Source

AI investment increase plans

80%+ of executives plan to boost AI investments in 2025 (even as other IT spending is scrutinized).

1

Current & planned adoption

29% of organizations already use agentic AI, and 44% plan to within a year – only ~2% aren’t considering it.

2

Market size growth

Global AI agents market is ~$7.9 B in 2025 and projected to reach ~$236 B by 2034 (~45.8% CAGR).

3

Startup focus on AI agents

Nearly 50% of Y Combinator’s Spring 2025 startups (67 of 144) are building AI agents or tools for creating them.

4

Productivity gains from agents

Using gen AI customer service agents increased issue resolution rates by ~14% per hour (with ~9% less handling time).

5

Adoption of AI agent tools

Microsoft reports 15 million developers using GitHub Copilot, and 230,000+ organizations (90% of Fortune 500) have used Copilot Studio to build AI agents.

6

Interest in agentic AI

52% of enterprises surveyed rank “GenAI for automation (agentic AI)” among their top generative AI priorities.

7

Generative AI adoption jump

The use of generative AI jumped from 55% to 75% of companies between 2023 and 2024, illustrating rapid uptake.

8

References: 1, 2, 3, 4, 5, 6, 7, 8

Explained: Each source above highlights a distinct trend in AI agent usage as of mid/late 2025, ranging from surging enterprise investment and adoption to burgeoning market growth, startup activity, productivity benefits to the mainstreaming of AI-agent tools. These surveys and reports collectively signal that AI agents will become deeply integrated across business functions and industries in H2 2025.

Section 3: H2 2025 – Emergence in the Wild

Now that we’ve seen the trends suggesting H2 2025 will be a big time for AI agents, it appears that the second half of 2025 is shaping up to be the moment when AI agents truly step out of the incubator and into more real business world use cases. This means businesses will start seeing practical, tangible impacts from autonomous AI agents across operations.

As such, for those businesses already implementing AI, you will now see the results. For those who haven’t implemented AI agents, now isn’t too late to catch up.

In the next three subsections, we’ll explore how these agents are evolving beyond text chat into multimodal capabilities, how companies are deploying them in continuous workflows with memory and self-improvement, and how a competitive marketplace for agents (and even entire “agentic” operating systems like CAIS’ agenticOS) is brewing among tech giants. 2025 has often been heralded as “the year of the AI agent,” and indeed the focus of H2 2025 will continue to shift toward these autonomous assistants as the big driver of efficiency and innovation in the workplace. Here’s what decision-makers need to know:

Section 3.1: Agents Go Multimodal

AI agents are breaking out of the text box and gaining native voice, vision, and even spatial understanding. In practical terms, this means an agent might converse with you…read more here.

Ross W. Green (created July 8, 2025). “Multimodal Agents in Action.” Canva.com.

Section 3.2: The Rise of Agentic Workflows

Early AI agents were often single-use bots for isolated tasks. Now we’re entering the era of agentic workflows and the agenticOS, whereby agents are being woven into ongoing business processes, carrying context over time and coordinating actions with minimal human prompting. Rather than a human handing off one task at a time…read more here.

Ross W. Green (created July 8, 2025). “On AI Memory.” Canva.com.

Section 3.3: Agent Marketplaces and “OS Wars

As AI agents proliferate, we’re also witnessing the birth of an agent economy, complete with marketplaces and platform battles reminiscent of the early mobile “app store” wars. Just as smartphones had ecosystems (iOS vs Android) where apps were bought and sold, the coming years will…read more here.

Ross W. Green (created July 8, 2025). “The 3 Agentic War Components Leaders Should Be Aware Of. Napkin.ai.

Section 4: 2026 – The Year of Autonomy

As AI agents mature past the experimental stage and we move out of H2 2025, 2026 is poised to be the year businesses grapple with true autonomy. In this section, we examine two emerging frontiers. First, we look at how some companies are restructuring teams around AI, even by piloting agent-led “pods” and largely synthetic teams. Next, we delve into the debate over AI agent identity and rights, from persistent digital personas to bold experiments granting agents legal status or decision-making authority. Decision-makers will see both practical trials and provocative questions as AI agents take on more human-like roles in 2026 and beyond.

Section 4.1: Agent-Led Teams & Org Restructuring

In 2026, forward-thinking organizations are embedding AI agents as integral team members, and in some cases, team leaders. The old static org chart is bending into what Microsoft researchers dub a dynamic “work chart,” built around goals and tasks rather than fixed departments…read more here.

Section 4.2: AI Agent Identity and Rights

As AI agents become woven into the fabric of work, an intriguing question arises: Should these autonomous agents have identities or rights akin to humans? In 2026, what was once a philosophical thought experiment is inching toward real-world tests. Companies and lawmakers are starting to consider…read more here.

Ross W. Green (created July 8, 2025). “Introducing: Portia.” Canva.com.

Final Thoughts: The Next Leap Has Begun

The data from 2025 makes one thing unmistakably clear: AI agents have moved beyond novelty and are now infrastructure. They’re now executing, learning, adapting, and in some cases, managing.

Enterprises large and small are embedding agents into daily operations, workflows are becoming autonomous, and entire markets are coalescing around this new paradigm. The second half of 2025 marks a transition point, but 2026 will test our readiness for true autonomy at scale.

From multimodal capabilities to agent-led teams and emerging legal frameworks for AI identity, the conversation is operational and no longer a theoretical debate within labs or amongst hungry startups. Simply put, as decision-makers, builders, and strategists, we’re seeing a reshaping of the contours of the intelligent workplace.

So, the real question isn’t “Should we use AI agents?” It’s: Are you designing your company for an agentic future, or are you watching it pass you by?

Stay tuned, because the agenticOS is only just booting up and 2026 will be here before we all know it!

Other resources:

2) Join our Community to access support from peers, a message board, and some great VIP content like our agentAcademy, weekly office hours, etc.

3) Join our weekly webinar series, "The Agentic Future with Devin Kearns" every Wednesday from 1-2 PM CST. Subscribe to this calendar for reminders.

4) Follow us on LinkedIn: Ross Green, CAiS, Devin Kearns

5) Want to learn more about how we work (e.g., build-with-you vs. build-for-you; Prebuilt SuperAgents vs, Customized Agents; etc)? Click here to schedule a meeting with us.

6) Have a friend who wants to sign-up for our Newsletter? Click here.

Reply

or to participate.