Agentic AI will see machines talking directly to machines to undertake shopping on behalf of consumers and B2B buyers. © Count Chris.
AI agents are cropping up everywhere. From virtual customer service bots to marketing automation tools, tech innovators are racing to launch solutions that promise faster service, smarter decisions and greater adaptability.
“AI agents are poised to become part of everyday life. Google’s Gemini helps plan your week, while OpenAI’s voice assistants manage tasks through natural conversation,” says Jonathan Reeve, Vice President, APAC, Eagle Eye.
“A wave of startups and innovators are already building AI agent solutions for specific business needs using foundation models from leading providers.”
Will the real Agentic AI please stand up?
Once limited to their own data, agentic AI models can now tap into broader data and capabilities via APIs and protocols such as Model Context Protocol (MCP), creating reliable links to external sources.
But with this evolution has come confusion.
Experts are calling on the industry to define agentic AI more clearly, to help business leaders cut through the hype.
“Let’s define it clearly: an agentic AI is one that acts like a true digital agent – setting goals, making decisions, taking action and learning independently,” says Marty Hungerford, Chief Innovation Officer at BRX.
So, what should a true agentic AI be able to do? According to Hungerford, it must:
1. Understand and interpret high-level goals
2. Plan and decompose tasks independently
3. Make decisions under uncertainty
4. Take action across systems without supervision
5. Self-correct and learn from feedback
“Recently, we built a zero-touch solution that scours the web for the latest AI news, writes a two-person script, generates two AI avatars to perform it, edits the video, and emails it for human review,” he explains.
“All of this happens without intervention. Impressive, yes. But agentic? No. It’s a finely orchestrated automated process linking multiple AI tools, not a self-driven agent. In fact, I’ve yet to encounter any AI that ticks all the boxes above.”
AI assistants vs AI agents
There’s a subtle but critical difference between AI assistants and agents, though they’re often used interchangeably.
AI assistants help users with tasks and queries, requiring ongoing input. AI agents, however, are more autonomous – capable of operating and making decisions with minimal or no human direction.
A strong example of an AI assistant comes from Nexxen, a global advertising tech platform. Earlier this year, the company launched nexAI – the integration of generative AI into their Nexxen Data Platform. This included a UI assistant within Nexxen Discovery, its proprietary insights tool.
With nexAI, clients can rapidly turn complex consumer data into actionable audience profiles and campaign plans.
“Our clients are continuing to lean into data and technology to navigate the fragmented media landscape, and nexAI meets this evolving need,” says Karim Rayes, Chief Product Officer at Nexxen.
“By integrating AI across our unified platform, and leveraging our existing data to inform these capabilities, we’re not just adding features; we’re fundamentally transforming the way campaigns are run and inventory is monetised.”
‘Conversational commerce’ vs AI agents
Another buzzword often mentioned in the same breath as agentic AI is conversational commerce. While related, they serve different purposes. Conversational commerce uses AI-powered chatbots and messaging tools to enhance customer engagement and online shopping. Agentic AI, in contrast, involves more advanced systems that can make autonomous decisions, solve problems and manage workflows.
“Conversational commerce is not just changing how consumers shop; it’s transforming how creative producers build brand experiences,” says Sangeeta Mudnal, Chief Technology Officer of GenAI platform Glu.
“As AI assistants increasingly mediate the relationship between brands and consumers, we’re witnessing a profound shift in how creative work is conceptualised, produced, optimised, and delivered. As an example, Glu isn’t simply another productivity tool but rather the foundation for the evolution in conversational commerce.”
Agentic AI and customer interactions
The intersection of AI, retail and marketing is rich with opportunity. Eagle Eye, for instance, has an advanced personalisation engine and predictive systems designed to ingest and process data intelligently.
As AI agents become smarter, they won’t just respond to queries, they’ll make decisions, compare options and even steer users to better offers. This will have a major impact on how retailers reach customers.
“Consider this scenario: a customer asks their AI assistant, ‘Where can I unlock behind-the-scenes content as a member?’ If your program’s benefits can’t be found and understood by that assistant, you’ll be excluded from consideration,” says Reeve.
“AI agents, personal shoppers and deal-hunting assistants will change how brands promote their products and offers. The way large language models and agents process information will likely lead to a reorganisation of marketing strategies and loyalty structures.”
Sarah Richardson, Founder and Director of The Australian Loyalty Association (ALA), agrees. She believes agentic AI is a game-changer for loyalty and personalisation.
“Agentic AI will be most transformative to the loyalty landscape,” she says. “Having an agent that can answer all your queries with relation to your membership as well as past purchase information helps brands to get on the front foot with customer expectations.
“Emerging technologies such as voice assistants and visual search are also creating new pathways into loyalty ecosystems, so there’s plenty of innovation that AI will bring!”
Billy Loizou, APAC Area Vice President at Amperity, sees a similar transformation unfolding in how brands compete.
“Imagine a world where your next purchase isn’t selected solely by you, but by an AI agent acting as your personal shopper,” he says. “Need an autumn outfit? Your AI agent instantly scours online stores, considering your size, style preferences, budget, event theme, and even the weather forecast to deliver perfectly tailored recommendations.”
‘No Hype’ agentic AI for real-world use
While many AI agents remain theoretical or experimental, some companies are quietly building practical, business-ready solutions.
One example is Pronto Software. The Australian enterprise resource planning and analytics provider has signed a strategic agreement with IBM Australia, integrating IBM’s Watsonx capabilities into the Pronto Xi ERP platform.
“We’re using AI to elevate workers, not replace them,” says Chad Gates, Managing Director at Pronto Software. “Our customers, many of them family-run, mid-sized businesses, can enable staff to act strategically. Pronto Software can work with customers to build and deploy agentic AI that not only informs, but acts on the information, unlocking real business value without compromising security.”
Another compelling example is RedOwl, a new platform reimagining business transactional workflows through AI and automation.
“AI agents are revolutionising how the modern enterprise operates,” says Jitto Arulampalam, Chief Executive Officer at RedOwl.
“At RedOwl, we have seen AI’s ability to operationalise board mandated governance, compliance and control across the organisation. We also see a future where AI agents are delivering board managed governance and control in real time.”
Are Aussie brands ready for agentic AI?
Despite the hype, the reality is that most businesses – especially in Australia – aren’t ready for true AI transformation.
The recent Digital, Marketing & eComm inFocus 2025 report revealed that 75 per cent of surveyed brands felt their eCommerce maturity lags. Only 2.5 per cent believed they were on par with global leaders.
Teresa Sperti, Founder and Director at ArkticFox, says that with digital ‘houses out of order’, many brands may struggle even more as AI disrupts traditional shopping experiences.
“Agentic AI will see machines talking directly to machines to undertake shopping on behalf of consumers and B2B buyers,” she says. “That will completely up-end the shopper journey as we know it, as it means we need to market as much to the machines as we do to shoppers.
“I believe retailers who don’t understand where the industry is headed are at risk of extinction within five to ten years, given that the vast majority of product discovery for most categories now starts online.”
Anthony Cipolla, AI Lead at COSOL, sees similar gaps across asset-centric industries. While organisations are interested in AI, many need help navigating the complexity.
“Verticals that rely on Enterprise Asset Management (EAM) are undergoing a revolution whereby traditional high manual effort required by humans to establish and maintain quality digital twins and master data will be rapidly replaced with semi-to-fully-autonomous agents,” he says.
“AI is both a disruptor and an enabler, and no doubt there will be tensions and hurdles along the way. Cultural change, mindset and trust will be key factors that organisations either have faced, are facing or will face along their efforts to modernise with data, AI and automation.”