Artificial intelligence continues to evolve, bringing transformative technologies that reshape the workplace. Among the most exciting advancements is Agentic AI, a step beyond traditional AI systems. Unlike generative AI models like ChatGPT, which excel at creating content, agentic AI introduces a new paradigm: proactiveness and autonomy.
As workplaces grapple with increasing complexity and rapid technological advancements, agentic AI offers a path to streamline operations, enhance decision-making, and spark innovation. This next frontier in AI goes beyond assisting humans — it proactively solves problems and drives results.
In episode 439 of our For Immediate Release podcast, published on 13 December, Shel Holtz and I explored this emerging trend and highlighted its transformative potential for 2025. As Shel aptly put it, “Agentic AI might be the most dramatic change computing brings to work and life since the introduction of the PC.”
What Is Agentic AI?
In a Harvard Business Review article on 12 December, author Mark Purdy describes agentic AI as representing a new frontier in artificial intelligence, characterised by its ability to act autonomously to achieve goals.
This means it doesn’t require constant human guidance. Instead, it leverages advanced technologies — machine learning, natural language processing, and automation — to optimise objectives such as improving efficiency, increasing sales, or enhancing customer satisfaction.
In our podcast discussion, Shel described scenarios where agentic AI takes centre stage: “Imagine a research agent analysing 100,000 tweets per hour, uncovering unmet needs and feature requests, or a content agent optimising 200 posts daily across platforms like X, LinkedIn, and TikTok.”
How Does Agentic AI Differ from Generative AI?
While both agentic and generative AI systems rely on similar underlying technologies, their applications differ significantly:
- Focus: Generative AI is about content creation (e.g., generating text or images), whereas agentic AI focuses on decision-making and action.
- Operation: Generative AI depends on user prompts to initiate activity, while agentic AI operates autonomously, guided by its objectives.
- Capabilities: Agentic AI can execute complex sequences of actions, such as searching databases, optimising supply chains, or triggering workflows, without direct human input.
Why Is Agentic AI Important for the Workplace?
Agentic AI offers several key benefits that promise to transform industries:
- Enhanced Specialisation: It enables greater workforce specialisation by automating granular tasks. For example, AI agents can handle information retrieval, workflow generation, and employee support, all while coordinating with other agents to optimise outcomes.
- Improved Innovation: With advanced reasoning capabilities, agentic AI can experiment and innovate faster than traditional methods. Multi-agent systems have already been used in scientific research, identifying novel materials and compounds in record time.
- Greater Trustworthiness: By reducing common errors like hallucinations (invented outputs in generative AI), agentic AI ensures more reliable decision-making. Its ability to assess and verify information sources boosts confidence in its actions.
Shel also pointed out how agentic AI could dramatically reduce operational costs: “What previously required 20 people and two million in salary now happens automatically with 2,000 in agent costs.”
Real-World Applications of Agentic AI
Google’s “entry into the agentic era” highlights the growing momentum behind this transformative technology. On 11 December, The Verge reported that Google introduced Project Mariner, a research prototype exploring new capabilities for agentic AI, alongside updates to its Project Astra.
Along with Gemini 2.0, these initiatives demonstrate Google’s commitment to advancing autonomous systems that can reason, adapt, and execute complex tasks. This push further solidifies the potential for agentic AI to reshape industries and workflows.
Agentic AI is already making waves in various fields:
- Customer Service: Dynamic AI agents can predict customer needs, resolve queries, and even offer proactive solutions — far surpassing traditional chatbots.
- Healthcare: Virtual caregivers powered by agentic AI can assist patients with empathy and efficiency, organising appointments, providing reminders, and offering companionship.
- Supply Chain Management: Agentic AI systems dynamically adjust logistics in response to real-time events like geopolitical disruptions or weather changes, effortlessly optimising operations.
On the podcast, we highlighted examples from Google’s top digital marketing trends and predictions for 2025, such as its Mariner and Astra projects simplifying workflows for companies like Kingfisher and enabling seamless integration of enterprise data. Cases like this demonstrate the growing accessibility of agentic AI, even for organisations that do not overhaul existing systems.
Guillaume Roques, Google’s Senior Director of Marketing for Google Cloud EMEA, describes such tools as “Chief Simplifier Officers,” representing a transformative advancement in 2025, building on the foundation laid by large language models in 2024.
Challenges and Opportunities
Despite its potential, agentic AI is not without challenges. Organisations must establish clear goals and create robust frameworks to integrate these systems effectively. Trust and oversight remain critical, as even the most advanced AI can make mistakes (such as what we know as “hallucinations” in generative AI).
To fully harness the power of agentic AI, businesses need to focus on:
- Setting SMART Goals: Ensuring objectives are specific, measurable, achievable, relevant, and time-bound.
- Effective Team Design: Balancing human and AI roles to achieve harmony and efficiency.
- Scaffolded Decision-Making: Creating safe and supportive environments for AI to learn and operate effectively.
As Shel observed, “Governance and security will be crucial for adoption. While the transformation won’t happen overnight, the benefits are already within reach.”
Embracing the Future
In his Harvard Business Review article, Mark Purdy highlights that early adoption and careful governance are critical to maximising the benefits of agentic AI while mitigating potential risks. As businesses move into 2025, leveraging these technologies thoughtfully will pave the way for significant advancements.
“The spread of AI agents will accelerate massively in 2025 as they become more accessible and easier to build. In the past, you had to change your information system to adopt technology, but AI agents put autonomous intelligence at your fingertips, empowering you to seamlessly integrate AI into your operations.”
Guillaume Roques, Google
Agentic AI is set to redefine the future of work. By combining the creativity of generative AI with the proactive capabilities of agentic systems, businesses can unlock new levels of innovation and efficiency. However, careful planning and ethical considerations are essential to ensure these technologies are deployed responsibly.
How ready are you — and your workplace — to embrace its transformative potential?
Listen to FIR 439
You can listen to or download our 20-minute conversation right here; or, if you don’t see the embedded audio player below, listen on the episode 439 show notes page on the podcast website.
You can find links there to the source material we used in this episode, along with the verbatim transcript of our conversation.
Related Reading & Listening:
- FIR #439: Agentic AI Tops Digital Trends for 2025 (FIR website, 13 December 2024)
- Perspectives on Industry Trends and Implications for Communicators (29 October 2024)
- NotebookLM Drives AI Audio Conversations That Feel Human (26 September 2024)
- Understanding AI in the Context of History (10 May 2024)
Footnote:
The generative AI hero image at the top was created by DALL-E 3 via ChatGPT Plus. I gave ChatGPT Plus this simple prompt:
“Create this image in landscape mode: A research agent analysing 100,000 social media posts per hour, uncovering unmet needs and feature requests, with a content agent optimising 200 posts daily across platforms like X, LinkedIn, Bluesky, and TikTok.”
In turn, the AI gave DALLE-3 this expanded prompt:
“A photorealistic image in landscape mode of a futuristic AI-powered office where a research agent AI is analysing 100,000 social media posts per hour. The scene features a high-tech environment with holographic displays showing social media feeds, data analytics, and charts. In the same space, a content agent AI is optimising 200 posts daily, with projections of posts being tailored for platforms like X, LinkedIn, Bluesky, and TikTok. The workspace is dynamic and collaborative, blending cutting-edge technology with sleek, minimalist office design. The atmosphere is vibrant, showcasing the potential of AI in streamlining tasks and uncovering insights.”
