The future is autonomous: Understanding agentic AI and its potential for operational support services

Just as we bring ourselves back up for air from the breathtaking impact of generative AI, a new frontier is emerging: Agentic AI. This next wave of AI, also known as AI agents, represent a significant leap forward, promising to transform how businesses operate by enabling AI to not just generate content, but to autonomously perform tasks and achieve goals. Understanding the potential of Agentic AI is now critical for future-proofing business strategies, including essential operational support functions for financial services, law firms, and professional services firms.

This isn’t just about enhanced automation; it’s about introducing a new level of intelligent autonomy into workflows. Gartner predicts that, “by 2029, agentic AI will autonomously resolve 80% of common customer service issues without human intervention, leading to a 30% reduction in operational costs.”[1] While this statistic focuses on customer service, the underlying capability—AI agents taking decisive action—has far-reaching implications across all sectors. As this technology matures, it’s crucial to approach its potential with a balanced, informed, and strategic perspective.

What is Agentic AI? Beyond content generation

So, what exactly sets Agentic AI apart? Generative AI, like the technology underpinning tools such as ChatGPT, Microsoft CoPilot, and Google Gemini, excels at creating new content—text, images, audio, and code—based on patterns in its training data. You instruct it, and it provides an output. Agentic AI or an AI agent, however, takes this a step further by making decisions, and acting autonomously to achieve specific goals. These systems can be given a high-level objective and will independently devise and execute the steps needed to accomplish it, often by interacting with external software tools. Think of it as moving from a skilled writer (generative AI) to a proactive personal assistant or a project manager (Agentic AI).

Key differences and technical considerations

The fundamental difference lies in the level of autonomy and proactivity.

• Generative AI: Requires explicit human instruction for each step and is primarily focused on content output.
• Agentic AI: Can operate with minimal human input once a goal is set, capable of multi-step reasoning, problem-solving, and task execution. It can query multiple large language models (LLMs) and other data sources, synthesize the information, and then decide on a course of action.

This capability to interact with multiple systems and make decisions is a key technical differentiator. For instance, an AI agent could be tasked with planning a client meeting. It might consult calendar applications, book travel through an airline’s system, arrange catering via a vendor platform, and then send out invites – all without any human intervention.

The potential capabilities of Agentic AI in support of business processes are vast, particularly in streamlining day-to-day tasks and administrative work, making life easier for professionals across various sectors.

Potential capabilities

• Automated workflows: Managing complex, multi-step processes like client or matter onboarding, due diligence, or project management
• Personalized assistance: Acting as a sophisticated digital assistant that can manage emails, schedule meetings, book travel, and even anticipate meeting or travel needs
• Data analysis and reporting: Gathering data from diverse sources, analyzing it, and generating comprehensive reports or presentations
• Research and information synthesis: Conducting in-depth research on complex topics, summarizing findings, and even drafting initial documents

Potential impact to knowledge workers

• Moody’s has reported that users of an AI research assistant (a type of agentic tool) consumed 60% more research while reducing task completion times by 30%.[2]
• A study by Deloitte predicts that AI could automate up to 100,000 legal jobs by 2036.[3]
• By 2030, AI could take over nearly 40% of tasks traditionally performed by lawyers, highlighting its potential to transform the industry.[4]

Potential use cases

• Financial services (investment banks, asset management firms, private equity)
  • A banker or analyst working on a pitch could delegate tasks like market research, competitor analysis, and initial presentation drafting to an AI agent. The agent could pull financial data, create charts, and assemble a preliminary deck, saving hours of manual work
• Law firms
  • A lawyer needing administrative support could use an AI agent for document processing, summarization of case files, legal research, e-discovery, and even scheduling depositions or client meetings
• Other potential use cases
  • Automating compliance checks, fraud detection, and know-your-customer inquiries
  • For operational leaders such as COOs, CAOs, transformation leads or heads of change management and automation, Agentic AI could identify process inefficiencies and even propose or implement automated solutions
  • Enhanced project management through AI agents that track progress, manage resources, and flag potential issues

Limitations and considerations

Despite the immense potential, there are limitations and critical considerations:

• Data dependency and access: AI agents are only as good as the data they can access. Ensuring secure and appropriate data access is paramount
• Complexity of implementation: Developing and integrating robust Agentic AI solutions requires significant investment and expertise
• Accuracy and reliability: Ensuring the accuracy of AI-driven decisions and actions is crucial, especially in high-stakes environments
• Humans in the loop: While autonomous, human oversight will still be necessary, particularly for high-risk tasks, to ensure ethical and responsible use, and to intervene when necessary. The EU AI Act, for example, emphasizes human oversight for high-risk AI systems
• Governance and trust: As these systems become more autonomous, establishing robust governance frameworks and building trust in their decision-making capabilities will be key challenges. Questions around accountability, transparency, and how to manage AI agents that might operate indefinitely need to be addressed
• Avoiding overhype: It’s vital to distinguish between the current, demonstrable capabilities of Agentic AI and its future, aspirational potential. The technology is still maturing

Embracing innovation for the future of operational support

Agentic AI represents more than an incremental technological improvement; it signals a fundamental step-change in how businesses will operate and how human expertise will collaborate with intelligent systems. For operational leaders in business, the immediate future is about strategic education, exploration, and preparing for an environment where autonomous AI agents become indispensable partners in driving efficiency, innovation, and significant value.

The journey towards fully realizing the potential of Agentic AI will undoubtedly involve navigating technical complexities and evolving ethical considerations. However, the promise of creating more streamlined, productive, and ultimately more human-centric work environments is compelling. At Williams Lea, we are excited to be an integral part of this evolving landscape. We are committed to continuously seeking and developing innovative ways to support our clients, helping them harness the power of an autonomous future and redefine what’s possible to support our clients’ critical business functions.

 Learn more how our tech-enabled presentations services are combining expertise and agentic AI into critical presentation activities and helping bankers, consultants and lawyers boost productivity.

[1] Source: Gartner, “Agentic AI Set to Transform Customer Service & Support Landscape“, March 5, 2025

[2] Source: Moody’s, “The rise of agentic AI in financial services: from automation to autonomy”, March 4, 2025.

[3] Source: A3Logics, Deloitte, “Introduction To AI Agents In The Legal Sector”, February, 13 2025.

[4] Source: A3Logics, Deloitte, “Introduction To AI Agents In The Legal Sector”, February, 13 2025.