The dawn of Agents as a Service: Where the market stands as we near the midpoint of 2026

As we get deeper into 2026, the conversation in corporate America is rapidly shifting from one-off uses of AI-powered chatbots into a more comprehensive and transformative era: the Agentic age.

AI agents, capable of performing complex, multi-step tasks autonomously, are already taking center stage, as companies across the globe look to transform business models and generate tangible ROI from AI after years of experimentation with the technology.

The industry perspective is clear. NVIDIA CEO Jensen Huang has declared that the AI industry has reached an inflection point, shifting from generative AI (Gen AI) chatbots to "Agents as a Service" (AaaS). Huang asserts that every SaaS company will evolve into an AaaS company, where AI agents act as digital employees – working, researching and executing tasks on behalf of users rather than focusing mainly on text generation.

From our perspective as venture investors, here is what we are seeing as the key trends to keep an eye on.

1. Shifting from copilot to autonomous agents

The AI revolution continues to accelerate at an unprecedented pace. The initial explosion of coding assistants has paved the way for a more organized market structure. We are now entering a "channeling" phase where major cloud providers are launching agent marketplaces, and enterprises are learning to adopt these systems at scale.

Looking ahead, the industry is poised for a significant shift from human-assisted copilots to truly autonomous agents. These systems will not just wait around and wait for a prompt but instead act 24/7 as independent digital co-workers, capable of orchestrating numerous specialized models to achieve complex tasks. A key enabler for this will be the rise of "Reinforcement Learning as a Service" (RLaaS), allowing agents to learn and improve continuously, creating efficiences for enterprises over time.

2. The future of Services-as-Software (SaaS)

Agentic AI is fundamentally reshaping the economics of software and services. While the AI funding landscape is seeing record highs, with capital heavily concentrated in foundational model players, the traditional SaaS model is under significant pressure.

We are witnessing the emergence of "Services-as-Software," where AI-native startups are converting what was once human labor spend into software spend. In other words, instead of selling a tool or "seat", these companies are selling an outcome. This is clear across key shared corporate functions:

• IT: Moving toward AI-driven, automated incident resolution
• Sales & Marketing: Disintermediating legacy CRMs with AI layers that Sales interfaces with directly
• Legal: Shifting from lawyer-assisting copilots to firm-serving autopilots
• HR: Automating top-of-funnel recruitment tasks like screening and outreach

This transition implies a major shift in who is doing the work, from human to agent, that accompanies a change in pricing models — from seats to usage, and ultimately, to purely outcome-based models.

3. Specialized AI models

While massive frontier models capture headlines, the real enterprise value is increasingly found in a diverse ecosystem of specialized models designed for specific tasks. Three key categories are emerging:

• Large Tabular Models (LTMs): Purpose-built for the structured, relational data that powers banks' most critical functions like fraud detection and credit risk assessment. These models deliver auditable predictions without months of manual machine-learning work.
• Large Action Models (LAMs): Designed to plan and execute multi-step workflows across different software applications, bridging the gap between understanding user intent and acting on it.
• Small Language Models (SLMs): Compact, domain-tuned models that require less computational power and can be deployed on-premise or at the edge, ensuring data privacy and delivering high accuracy for specific tasks.

4. Security and Trust

The proliferation of AI agents introduces a complex new landscape of risks. Threat actors are using AI to generate sophisticated attacks with increasing speed, including compelling deepfakes and accelerated phishing campaigns. Internally, as AI agents inherit permissions and access sensitive data, they create "Shadow AI" attack surfaces that traditional security was never designed to handle.

This reality necessitates a fundamental rethink of security. Organizations must adapt by empowering their security teams with AI-native tools and comprehensive training. The defense strategy is evolving around three key pillars.

• First is the rise of the AI-powered Security Operations Center (SOC). Traditional SOCs, often short-staffed, have struggled with the volume of alerts. While earlier automation tools helped, their rules-based nature limits their ability to counter novel, AI-generated attacks. A new generation of Agentic AI SOC platforms, like that from Prophet Security, acts as a force multiplier. These platforms use Large Language Models (LLMs) to work alongside human analysts, automating investigations and accelerating threat response times from minutes to seconds.
• Second, proactive employee training is critical. With deepfake threats extending to voice, text and video, the human element is a key line of defense. Companies like Adaptive Security are addressing this with a three-pronged approach: interactive security awareness training, multi-channel phishing simulations (email, text, voice, and video), and automated risk scoring.
• Finally, securing the AI ecosystem itself requires new foundational controls. AI agents cannot be treated as mere extensions of human accounts, they need their own scoped, temporary and auditable identities to create a secure control layer. Agents will also need task‑scoped permissions, and real‑time policy enforcement—allowing agents to operate autonomously without inheriting dangerous, overscoped credentials.

The Agentic age is not a distant vision, it is already here. We will be monitoring all of the above factors and continue to weigh in on how we in venture capital can help companies navigate agentic AI's immense opportunities and profound challenges.