Agentic AI transforms business operations with enhanced oversight
The integration of agentic artificial intelligence (AI) into business operations is gaining significant momentum across industries, with new research, commentary, and product announcements underscoring both the promise and complexities of these advanced technologies.
Matt Johnson, Managing Director for AI & Data at Temus, outlined the evolving landscape of AI agents, noting an industry-wide shift from rudimentary AI interactions towards more advanced, contextually aware systems. "We're witnessing a significant shift in how AI agents are being deployed across industries. The most successful implementations go far beyond basic prompting," Johnson observed. He highlighted the application of sophisticated techniques such as automated reprompting, parameter-efficient fine-tuning, and reinforcement learning, which allow agents to learn from their environments and incorporate expert knowledge.
Johnson emphasised that data remains the critical foundation for agentic AI. He noted, "Companies are now realising they need deliberate strategies to acquire and structure this expert knowledge – it's become a competitive differentiator." In sectors such as healthcare and financial services, he asserted, the inclusion of human-in-the-loop workflows is not optional but essential, with the best AI systems augmenting human expertise rather than replacing it.
The software development sector, according to Johnson, has provided one of the most compelling success stories, with AI tools such as Claude Code assisting developers by providing contextual suggestions and even autonomously generating code, all while preserving human oversight. This reflects a broader trend, with organisations increasingly viewing AI agents not as autonomous replacements for professionals, but as tools to enhance productivity and decision-making.
In the domain of cybersecurity, a new study from Cycode, presented at the RSA Conference 2025, illuminated how agentic AI is reshaping application security practices. The survey found that while 60% of cybersecurity professionals remain in early stages of adoption, those organisations that have embraced agentic AI report notable productivity gains and reduced risks in development and security workflows.
Amir Kazemi, Director of Product Marketing at Cycode, observed, "Many interpretations and modalities of 'agent' exist, from simple chatbots to complex workflow automations to true autonomous agents. Our data underscores that educating the market on what agentic AI truly is, why it matters for AppSec, and its tangible value is paramount right now."
The Cycode research illustrated growing interest, with almost 50% of surveyed professionals planning to adopt agentic AI in the coming year. Yet, concerns remain about granting AI systems autonomy, with businesses taking a measured approach to integrating these tools. The study identified key opportunities: 44% of professionals believe agentic AI will improve vulnerability management, while 52% see significant value in using AI-driven security checks at the code commit stage. The perceived widening gap between application security and development resources, with some teams managing ratios as high as one security specialist per 1,000 developers, exemplifies the mounting pressure on teams that agentic AI could help alleviate.
Financial services are also experiencing AI-driven transformation, as demonstrated by the launch of GTreasury's GSmart AI platform, designed specifically for treasury and finance operations. The platform aims to deliver efficiencies and transparent insights for CFOs and treasury professionals facing complex market and regulatory conditions.
GTreasury CEO Renaat Ver Eecke stressed the necessity for AI in finance to prioritise security, compliance, and rapid problem-solving. "GSmart AI... empowers CFOs and treasury teams to confidently take advantage of powerful insights and value without sacrificing compliance or oversight," Ver Eecke stated. The platform provides automated analysis, risk identification, and strategic recommendations, all while ensuring auditability and governance. Mark Johnson, Chief Product Officer at GTreasury, added that GSmart AI is distinguished by its transparency and data sovereignty features, supporting rigorous standards and regulatory requirements.
These developments signal that agentic AI, when combined with robust data strategies and clear boundaries for human oversight, is rapidly becoming integral to modern workflows. Whether in software development, cybersecurity, or treasury operations, organisations are increasingly seeking to leverage the unique capabilities of these AI agents to enhance human judgement, streamline complex tasks, and maintain compliance in a rapidly evolving technological landscape.
