OPINION: In recent years, there has been a surge of excitement around artificial intelligence, particularly with the rise of generative AI, which can create text, images and even code based on prompts. However, a new evolution in AI is emerging – one that takes things a step further. This is what we refer to as Agentic AI. Rather than just generating content, Agentic AI acts on it, executing tasks and decisions that traditionally required human involvement. While it brings significant potential, it also requires careful consideration of the implications.

As the Co-Founder and CTO of Inde Technology, I’ve spent a great deal of time thinking about how Agentic AI can reshape industries, and how we, as businesses and individuals, can best navigate this shift. In September, I was invited to speak at the Tech Summit in Christchurch, alongside AWS Senior Solutions Architect, Casey Orr, to discuss ‘Agentic AI – A new frontier?’ It was great to connect with so many people keen to understand the pros and cons and what the exciting opportunities are that lie ahead with this new form of AI.

To put it simply, Agentic AI takes us from making stuff to doing stuff! It moves beyond passive content generation to active task execution and, unlike chatbots or traditional AI systems that merely respond to inputs, Agentic AI is designed to act autonomously on a series of tasks, freeing up time and resources for more strategic decision-making.

For example, while generative AI might help you write code for a homepage, Agentic AI would go further by building the entire website – writing the code, generating content, uploading it to a hosting provider, testing it and handing over a ready-to-use product. This ability to take autonomous action, rather than just providing information or outputs, is where Agentic AI truly comes into a league of its own.

Real-World Applications

The potential applications for Agentic AI are vast and varied. As well as the website development example above, there are opportunities such as:

•  Travel assistance: Rather than just using chatbots to explore flight and accommodation options, Agentic AI could become really useful and rebook your hotel and flight without requiring human input when travel plans change due to delays or cancellations.
• Healthcare: Digital avatars powered by AI could be of huge help to nursing, supporting on low-risk and non-diagnostic tasks. It could also support by transcribing patient symptoms, guiding and instructing patients on prescribed medication, as well as following up with doctors, automating the more repetitive but essential tasks and allowing healthcare providers to focus on more complex patient needs.
• Scientific research: Agentic AI can streamline the research process by combining various AI models, datasets and research papers. It chains together different modes of operation, offering experiments to try and process the outputs to suggest further actions. This enhances the speed and efficiency of scientific discoveries, meaning the medical cures for many problematic diseases could be just around the corner.
  
  

These examples only scratch the surface of what is possible with Agentic AI, but they also raise important questions about the future of automation and human roles.

Read more: GenAI in Design: Tool or Threat? Why Designers Must Evolve to Stay Ahead

The Pros and Cons

As with any powerful technology, Agentic AI brings both advantages and challenges. Understanding these helps us approach its integration with balance. I’ve summarised the key points discussed below:

The benefits:

• Efficiency and speed: Agentic AI can pull data from various sources and act on it in real-time, offering solutions faster than any human team could. This makes it incredibly valuable for customer service, logistics and operational efficiency.
• Task automation: AI takes over the repetitive tasks, allowing us to focus on more meaningful, innovative and strategic work. By handling lower-level decision-making, it enhances our capacity to make better, more informed choices.
• Improved accuracy: With its ability to access real-time data and automate tasks, Agentic AI can reduce the chances of human error in routine operations, making processes more reliable.
  
  

 The challenges:

• Erosion of problem-solving skills: There’s a concern that if AI handles the majority of routine and critical thinking tasks, humans might lose their ability to solve problems creatively. Over-reliance on AI could weaken our intellectual independence over time.
• Ethical challenges: In areas such as healthcare or defence, the autonomy of Agentic AI raises ethical concerns. For example, could we trust AI to make decisions about medication or military actions? The idea of allowing AI to make life-impacting decisions without human oversight can feel unsettling.
• Limited emotional intelligence: While Agentic AI excels at logical decision-making, it lacks the emotional intelligence that humans bring to the table. There’s a risk that in situations requiring empathy or cultural understanding, AI may fall short.
• Decisions in isolation: Agentic AI models generally operate in isolation, making decisions based on the model, logic and decision criteria – and any deliberate or inadvertent biases that may be present. They don’t operate in a society, making decisions based on many external factors like humans do. This can potentially lead to outcomes that lack the nuance or ethical consideration we expect in our complex human environments.
  
  

Navigating the risks – how to manage Agentic AI responsibly

While these challenges are real, they are not insurmountable. By managing the integration of Agentic AI thoughtfully, we can maximise its benefits while mitigating the risks.

• Human-AI collaboration: Instead of fully handing over decision-making to AI, we can create systems where AI works alongside humans. This allows AI to handle the tedious tasks while humans focus on critical thinking and emotional nuance, ensuring we maintain control in high-stakes decisions.
• Ethical guidelines: Establishing clear ethical boundaries for AI use, especially in sensitive areas like healthcare or defence, is essential. By maintaining human oversight and setting clear guidelines for where and how AI can act, we can mitigate potential risks.
• Focus on education: Ensuring that future generations develop critical thinking and problem-solving skills is vital. Alongside AI, we should emphasise human development so that people can continue to innovate, iterate and tackle complex issues that arise when AI faces limitations.
• Transparency and safety mechanisms: By implementing clear safety protocols, such as override mechanisms, and ensuring transparency in AI decision-making, organisations can build trust in AI systems while keeping accountability firmly in human hands.
  
  

So what’s next?

Agentic AI is more than a tool – it’s a step toward reshaping how we think about automation and intelligence. But with great power comes great responsibility, and our goal is to help you navigate this shift thoughtfully, ensuring that AI works with us, not for us, in the truest sense.

By embracing the future of AI while remaining mindful of its challenges, we believe we can unlock new levels of efficiency, creativity, and opportunity for businesses across all industries.