A subtle but potentially significant shift is occurring in artificial intelligence (AI). Machines are advancing beyond processing commands to undertaking tasks with more autonomy.
Google’s release of Gemini 2.0 may mark a turning point in this evolution, showcasing AI systems that can independently navigate complex tasks across multiple platforms. For example, Gemini 2.0 powers projects like Astra, a universal assistant for Android devices, and Mariner, an agent capable of autonomous web navigation.
These developments suggest the system could transform user interactions and task automation. For businesses, such advancements indicate possibilities for AI to impact operations ranging from warehouse management to customer service.
“Gemini 2.0 improves on previous AI systems by advancing the capabilities of autonomous decision-making through the integration of more sophisticated AI agents that leverage real-time data processing and adaptive learning models,” Prashant Kelker, chief strategy officer, partner and lead consulting sourcing and transformation – Americas, with global technology research and advisory firm ISG, told PYMNTS.
“As a result, enterprises will need to strengthen the cross-functional alignment between technology, business and compliance teams. As agentic AI goes into production, we are expecting cloud and edge computing capabilities to scale.”
The key innovation lies in Gemini 2.0’s ability to handle multistep processes with reduced human oversight. Unlike traditional AI that responds to specific prompts, this system aims to autonomously coordinate across platforms, potentially managing inventory or processing orders.
“Rather than completely redesigning their eCommerce systems, businesses will likely extend existing accessibility and structured data standards to create an ‘AI-enhanced HTML’ layer that sits between pure visual interfaces and full APIs,” Dev Nag, CEO of QueryPal, a support automation company, told PYMNTS.
A distinguishing feature of Gemini 2.0 is its unified approach to processing different types of information. While previous systems often required separate tools for handling text, images and audio, this new model reportedly integrates them—a development seen as critical for real-world applications where data often spans multiple formats.
For retailers and logistics companies, such advancements might manage supply chains, from tracking shipments to predicting inventory needs, while also handling customer interactions across multiple channels. Financial institutions might consider deploying it to enhance fraud detection systems, potentially allowing human analysts to focus on strategic tasks.
With further development, Gemini’s agentic approach could have massive usefulness for practical and inefficient consumer tasks, Kevin Green, COO of Hapax, an AI for the financial services industry, told PYMNTS. But, he said, there are a few important things to note.
“First, agentic is not a tidal wave. It will not come in and impact every aspect of consumer life. Much of what we do is already incredibly efficient and will not be improved upon, or at least not improved upon quickly, by agents.
“Take online shopping for example. While AI may allow us to more quickly find what we are looking for or better introduce us to new products, the act of purchasing on Amazon is already incredibly efficient so products like AI shoppers are not something I’m expecting to take hold.”
Gemini 2.0 could be an indicator of broader changes in business operations. The ability to process multiple data types and make decisions autonomously could have implications for retail and manufacturing industries.
Some early adopters are already exploring applications. For instance, logistics companies are testing AI agents for tasks like tracking shipments and rerouting them based on real-time conditions and customer preferences.
Similarly, Salesforce recently announced Agentforce 2.0, a platform designed to enhance sales, marketing and customer service through AI-driven solutions. Customer service departments are experimenting with agents capable of resolving complex support issues by accessing multiple systems without human intervention.
However, these advancements raise concerns. As AI systems gain more autonomy, ensuring their security becomes critical. A compromised AI agent could disrupt supply chains or execute unauthorized financial decisions.
The potential cost savings and efficiency gains are considerable but not guaranteed. By automating routine processes, companies might reduce operational overhead and improve response times, though the extent of these benefits remains to be seen.
For instance, SoundHound AI’s voice technology, used in automotive and restaurant sectors, highlights the interest in AI-driven solutions but also underscores the challenges in scaling such systems effectively.
For businesses, the outlook is complex. Autonomous AI is no longer just a theoretical prospect — it is beginning to influence how companies operate and compete. However, success will likely depend on balancing automation with appropriate human oversight. Companies starting small-scale implementations now may position themselves better as these technologies evolve.
We’re always on the lookout for opportunities to partner with innovators and disruptors.
Learn More