The integration of Agent Artificial Intelligence (AI) into business operations has gained great momentum across the industry thanks to new research, commentary and product launches that highlight both the promise and complexity of these advanced technologies.
Matt Johnson, managing director of AI & Data at Temus, outlined the evolving landscape of AI agents, focusing on the industry-wide shift from rudimentary AI interactions to more sophisticated and contextually recognized systems. “We are witnessing a major change in the way AI agents are deployed across the industry. The most successful implementations are well beyond the basic prompt,” Johnson said. He emphasized the application of sophisticated techniques such as auto-renewal, parameter-efficient fine-tuning, and reinforcement learning, allowing agents to learn from their environment and incorporate expertise.
Johnson emphasized that data continues to be an important foundation for agent AI. He says, “Companies now recognize that they need a deliberate strategy to acquire and build this expertise. It has become a competitive differentiator.” In sectors such as healthcare and financial services, they argued that including human workflows in loops was not an option and essential, and that AI systems were enhancing human expertise rather than replacing them.
According to Johnson, the software development department offers one of the most compelling success stories, with AI tools like Claude code helping developers by providing context suggestions and even generating code autonomously while maintaining human surveillance. This reflects a wider trend, with organizations increasingly watching AI agents not as an autonomous alternative to experts, but as a tool to enhance productivity and decision-making.
In the realm of cybersecurity, a new Cycode study published at RSA Conference 2025 highlighted how agent AI is reshaping application security practices. The study found that while 60% of cybersecurity experts remained in the early stages of adoption, organizations that adopted Agent AI have significantly increased productivity and reduced risks in development and security workflows.
“From simple chatbots to complex workflow automation to true autonomous agents, there are many interpretations and modalities of “agents” that exist,” says Amir Kazemi, director of product marketing at Cecode. Our data highlights what Agent AI is really, why it is important to AppSec, and its tangible value is correct.
The CYCODE study shows that almost 50% of the experts surveyed plan to employ Agent AI next year. However, concerns remain about granting autonomy in AI systems, and companies are taking a measured approach to integrating these tools. Important opportunities were identified in this study. While 44% of experts believe that agent AI will improve vulnerability management, 52% see great value when using AI-driven security checks during the code commit stage. The widened gap between application security and development resources manages the ratio compared to one security specialist per 1,000 developers, demonstrating pressure on teams that Agent AI could mitigate.
Financial Services is also experiencing AI-driven transformation, as demonstrated by the launch of Greasury’s GSMART AI platform. The platform aims to provide efficiency and transparent insights to CFOs and Treasury experts facing complex market and regulatory conditions.
Gtreasury CEO Renaat Ver Eicke highlighted the need for AI in finance to prioritize security, compliance and rapid problem resolution. “GSMART AI … will enable CFO and Treasury teams to confidently utilize strong insights and value without sacrificing compliance or oversight,” Ver Eecke said. The platform provides automated analysis, risk identification and strategic recommendations while ensuring auditability and governance. Gtreasury’s Chief Product Officer Mark Johnson added that GSMART AI is distinguished by its transparency and data sovereignty capabilities, and by supporting strict standards and regulatory requirements.
These developments show that agent AI, combined with robust data strategies and clear boundaries of human surveillance, is rapidly becoming essential to modern workflows. In software development, cybersecurity, or finance operations, organizations are increasingly hoping to leverage the unique capabilities of these AI agents to enhance human judgment, streamline complex tasks, and maintain compliance with the rapidly evolving technological landscape.
