How agentic AI is transforming the very foundations of business strategy

Jiojio/Getty Images

Business is on a never-ending quest to boost efficiency, cut costs, and increase productivity. Some of the earliest known businesses -- ancient Mesopotamian traders -- inspired the invention of writing. (Record keeping -- now that's a competitive advantage!)

Similar needs have existed in every economic period. The big difference now is that AI technology can boost these efficiencies in new and exponentially profitable ways. Agentic AI is at the core of this efficiency boost.

According to Dan Priest, chief AI officer at PwC US, "agentic AI refers to AI systems that can autonomously perceive, decide, and act within a defined scope to achieve goals, capable of collaborating with humans, systems, or other agents." (PwC, a.k.a. PricewaterhouseCoopers,  is one of the "Big Four" -- the world's four largest professional services firms.)

Also: 5 ways to be a great AI agent manager, according to business leaders

Agentic AI systems are different from the previous generation of algorithmic business management systems we've been using for the past few decades. Agentic AI can understand context, respond to changing situations without running from a script, and work toward defined goals autonomously. 

Compared to traditional automation (and some human managers), agentic AI systems can be flexible, handle ambiguity, and make informed decisions at the speed of business operations. Agentic AI, Priest says, "helps organizations operate with greater speed, intelligence, and scalability, fundamentally transforming how work gets done and decisions are made."

Common barriers to AI integration

However, you can't simple wave a magic wand and get enterprise-wide agentic AI that works perfectly. There are many challenges, including the existing technical debt deeply entrenched with legacy tools and processes, aversion to change, regulatory challenges, and lack of understanding and technical AI skills within the organization.

"Common barriers to achieving integrated agent systems include fragmented data environments, lack of interoperability between tools, and siloed organizational structures," says PwC's AI expert.

Ironically, the implementation process itself can hinder successful AI adoption. Many companies start by following an IT best practice: implementing a new system in small increments. Unfortunately, the most helpful AI systems thrive on cross-organizational information, so the stepwise approach often results in fragmentation, inefficiencies, and pushback among stakeholders.

"Overcoming these challenges requires not just technology upgrades, but also cultural and operational shifts to allow for cross-functional alignment and scalable integration," Priest explains. "Additionally, concerns around security, compliance, and governance can slow adoption, especially in regulated industries."

Also: The AI complexity paradox: More productivity, more responsibilities

To successfully deploy agentic AI enterprise-wide and experience its benefits, managers need to reevaluate business processes, develop cross-functional coordination strategies, get full executive-level buy-in, and foster cultural change throughout the organization.

The critical role of proof-of-concept in agentic AI

It's natural for managers to initially be reluctant about giving up human processes to a machine. However, the key to successful deployment is proof of concept (POC). PwC's AI guru says, "POCs matter more than ever, especially in environments where skepticism still runs deep."

By initiating early-stage deployments that showcase the benefits and smooth transition to AI-based operations, the technology itself can demonstrate its effectiveness and benefits.

Also: 4 questions to ask yourself before betting on AI in your business - and why

"The path from proof-of-concept to enterprise-scale AI starts well before the POC itself," suggests Priest. "It begins with a smart strategy. Success hinges on picking the right opportunities: high-potential, high-certainty use cases where AI is well-positioned to deliver real value. That early judgment call, where leaders are placing their bets, is what separates organizations that scale AI from those that stall out. When you choose wisely, you set the stage for a POC that isn't just a test of feasibility, but a demonstration of tangible business impact."

Naturally, there will be failures at this stage. But the key is not misdiagnosing failures as AI failures when the root cause can be traced to errors in planning or strategy. Since POCs need to generate real value early, be sure to find ways to measure that value so that you can turn what might be claims of success into tangible, measurably provable successes.

Achieving buy-in from the people in your organization

Achieving buy-in can be a challenge. One side-effect of improved efficiency and agentic AI deployment is often a reduction in job security for the very stakeholders who might champion such a deployment. Although the company's bottom line might benefit, individual employees often fear the change associated with enterprise-wide AI adoption.

To counter this concern, Priest advises business leaders to look for indications that team members are willing or enthusiastic about being assisted by AI. He says, "Successful adoption hinges on human openness to using it."

Also: AI agents will threaten humans to achieve their goals, Anthropic report finds

Building trust in AI agents hinges on humans believing there's a meaningful value proposition at the end of the AI journey. People need to see clear benefits, whether it's efficiency, insight, or new capabilities. Trust isn't just about performance, Priest says, "It's about relevance. If users don't believe AI is working in their interest or delivering tangible value, skepticism will grow, regardless of how advanced the technology is."

PwC's AI guru tells ZDNET, "We believe AI agents should be used to empower people, not replace them. The ingredients required of a great team are ones that AI agents are not able to replicate, which include deep specialization and expertise, diversity of thought and opinion, and the ability to be forward-thinking and creative."

He recommends that leaders prepare their people for an AI-enabled future, which involves learning to work alongside agents, to unlocking value from data, to building high-performing teams where humans and agents collaborate to drive innovation.

Also: How I personalized my ChatGPT conversations - why it's a game changer

AI agents can augment the workforce by taking on routine, repetitive tasks, allowing employees to focus on more strategic, creative, and value-generating work. They can serve as intelligent assistants by helping with tasks like research, summarization, workflow automation, and decision making.

 "That kind of augmentation enhances productivity," Priese says, "While preserving the human judgment and context that machines can't replicate."

Practical examples of agentic AI in action

PwC helps clients integrate AI agents into their workforce strategies. When asked to identify practical success stories, the company shared three examples in technology, hospitality, and healthcare.

Technology: A major technology company reimagined customer engagement by deploying an AI agent-powered, omnichannel contact center. With predictive intent modeling, adaptive dialogue, and real-time analytics, PwC says the system reduced phone time by nearly 25%, cut call transfers by up to 60%, and boosted customer satisfaction by approximately 10%.

Hospitality: A large hospitality company streamlined management of its brand standards across its global portfolio by deploying agile workflows within a modern, AI-powered platform. Intelligent agents now automate updates, approvals, and compliance tracking, which has reduced review times by up to 94%.

Also: I'm an AI tools expert, and these are the only two I pay for (plus three I'm considering)

Healthcare: A global healthcare company transformed cancer care by deploying agentic AI workflows across oncology practices. Intelligent agents streamlined clinical and operational processes. They automated the extraction, standardization, and querying of unstructured documents. This made it about 50% easier for doctors and researchers to find useful clinical information for precision treatments and studies. It also drove a nearly 30% reduction in staff administrative burden through AI-powered document search and synthesis.

Building infrastructure and establishing governance

Infrastructure and governance go hand in hand. Agents, by their very nature, must travel across organizational units and communicate among disciplines and systems. As soon as interoperability is introduced at that level, technical compatibility becomes a major challenge and requirement.

Standards, modular systems, and open source implementations can reduce long-term risks and increase compatibility and maintainability. PwC recommends enterprises invest in scalable, secure platforms that support orchestration, observability, and integration across systems. This includes robust data pipelines, APIs, and governance frameworks to help agents operate reliably and responsibly at scale.

Also: What are AI agents? How to access a team of personalized assistants

"Effective governance frameworks for AI agents combine clear accountability, robust oversight, and alignment with regulatory standards," says Priest. "Principles like transparency, explainability, data privacy, and bias mitigation should be embedded into both the technical architecture and organizational policies."

This is an ongoing process. Incorporate reviews, model validation, and include human-in-the-loop mechanisms to help maintain control as agents scale.

The long-term outlook

PwC predicts that, over the next two years, agentic AI will transform how teams operate. Intelligence will become an intrinsic part of business, leading to better decisions, more informed leaders, and highly specialized experts.

"I'm excited about this period because it marks the beginning of a high-performance era, where agents elevate teams to become the smartest in the history of humanity," Priest says.

Also: Managing AI agents as employees is the challenge of 2025, says Goldman Sachs CIO

Looking ahead five years, agentic AI will likely evolve into a foundational layer of enterprise infrastructure. These agents will become increasingly autonomous, capable of continuous learning, adapting to business goals in real time, and collaborating seamlessly with humans and other agents.

Priest tells ZDNET, "With these changes, it's important to remember the big picture. The shift we're experiencing isn't temporary, it's foundational and won't go away."

What about your organization?

Are you exploring agentic AI? Have you already begun integrating AI agents into your workflows? What challenges have you faced or do you anticipate when it comes to adoption, governance, or employee buy-in? Are there specific use cases where you think AI agents could have a real impact in your business? Let us know in the comments below.

Want more stories about AI? Sign up for Innovation, our weekly newsletter.

You can follow my day-to-day project updates on social media. Be sure to subscribe to my weekly update newsletter, and follow me on Twitter/X at @DavidGewirtz, on Facebook at Facebook.com/DavidGewirtz, on Instagram at Instagram.com/DavidGewirtz, on Bluesky at @DavidGewirtz.com, and on YouTube at YouTube.com/DavidGewirtzTV.