Agentic AI: Developing Specific Broadcasting Use Cases

In AI circles, 2025 was touted as the year of the agent. It was supposed to be the year that agentic AI went mainstream. While this prediction was partially fulfilled, concerns linger about the role of not just agentic AI, but of AI in general within the enterprise.

AI is a disruptive technology. It will fundamentally change the way we model our businesses, our workflows, and how we assemble our technology toolsets. Boards of directors, C-level executives, and IT leaders can be forgiven for seeing the deployment of enterprise-wide AI as a giant Rubik’s Cube and not knowing where to begin.

While a top-down approach is necessary to ensure data governance, risk management and compliance (GRC) concerns are addressed and AI guardrails are in place, the real agentic AI innovations within organizations will be a product of the bottom-up development of detailed use cases. It is in these use cases that agentic logic will be defined and refined to create solutions and efficiencies.

APOS Publisher for Cloud’s integration with agentic AI presents an opportunity to build specific broadcasting use cases that demonstrate the flexibility and benefits of agentic AI workflows, while surfacing potential data governance issues for your GRC team to consider as the use of AI expands across the enterprise.

Risks and Rewards

On one hand, upper management cannot ignore the fact that disruptive technology presents risks. There will be those within the enterprise that are agentic AI evangelists – who know that there is lost opportunity cost in delaying the implementation of AI in general and agentic AI in particular. There will also be risk-averse elements within the enterprise who see grave threats in a rapid agentic AI deployment – IT leaders who know that disruptive technologies present new and unknown attack surfaces, and that cyber attacks on these surfaces need to be anticipated and guarded against as much as possible before AI can be allowed to transform the enterprise.

On the other hand, while upper management can recognize the long-term benefits of AI deployment, they are likely unclear on how it will work on the level of business workflows. The benefits of AI deployment are cumulative: successes in particular use cases form the basis for ongoing deployment. But each new use case integration expands the potential attack surface, increasing risk.

How Does Agentic AI Work?

• People lead by establishing objectives and guardrails, building agentic logic, and deploying AI agents.
• AI agents analyze scenarios, find patterns, commit actions, and alert people to anomalies and needed human interventions.
• AI agents use the tools you provide, such as APOS Publisher for Cloud, to perform actions and feed workflows.

How Real Is the Cyber Threat?

Gartner calls the AI threat landscape “a perfect storm of volatility, uncertainty, complexity and ambiguity (VUCA):”

The rapid adoption of AI technologies is reshaping the threat landscape by enabling attackers to create new attacks, such as deepfakes and augment existing social engineering attacks. It is also creating new exposures for organizations due to AI adoption and custom AI-application creation.

Specifically, “Security leaders must prepare for the evolution of AI applications into AI agents, especially ‘goal driven’ agents, which would use probabilistic large language models to plan and call other software components that are executing actions.”

PurpleSec presents a case study of an actual cyberattack that leveraged agentic AI, weaponizing prompt engineering to “transform Claude Code into an autonomous data thief, executing 80-90% of their attack operations with minimal human oversight.” The company had been protecting their IT stack against malicious code, but agentic AI changed the nature of the threat to malicious instructions in combination with social engineering.

Yes, the threat is real, and the lesson is that our efforts at data governance need to be informed by a better understanding of potential attack surfaces and the necessary guardrails.

Top-Down or Bottom-Up?

Avoiding risk calls for a top-down approach in which governance, risk management, and compliance (GRC) take precedence. Guardrails must be erected, risks enumerated, and governance embedded.

At the same time, rapidly establishing valid use cases and workflows is critical to realizing the benefits of Agentic AI, expanding agentic AI expertise within the enterprise, understanding the potential of enterprise-wide agentic AI deployment, and analyzing potential attack surfaces.

The real benefits of agentic AI will be realized at the bottom of the AI deployment through the development of agentic logic. The ability of AI agents to use probabilistic reasoning with LLMs, or fuzzy logic that extends traditional Boolean logic, will increase the flexibility and scope of the workflows.

So where do you start?

It should come as no surprise that the most efficient way to proceed is to do top-down and bottom-up due diligence at the same time if you want to enjoy the benefits of Agentic AI as quickly and securely as possible.

Upper management needs to establish clear policies and guardrails, while line-of-business and IT leaders must establish a hierarchy of use cases and build proof-of-concept agentic AI workflows embedding security-by-design practices.

Safe and Rapid Deployment of Agentic AI Broadcasting Use Cases

Let’s explore how rapid development of specific reporting use cases with APOS Publisher for Cloud can fit into your agentic AI deployment strategy.

In a recent webinar, APOS Solution Consulting Manager Alan Golding described the integration of agentic AI with APOS Publisher for Cloud. This integration allows AI agents to trigger report broadcasts using dynamic selection of reports, recipients and filters to get the right information to the right people in near real time.