Industry executives and experts share their predictions for 2026. Read them in this 18th annual VMblog.com series exclusive.
By PwC leaders
In 2026, the AI story will shift from experimentation to execution. After years of pilots, proofs of concept, and incremental efficiency gains, businesses finally have something they’ve been missing: proof points. Clear benchmarks, working demos, and the rise of agentic AI – paired with stronger governance and engineering practices – are showing what it really looks like to transform AI at scale.
This year’s predictions signal a new phase of maturity. Businesses are moving past scattered use cases and toward focused AI programs that reshape workflows, reinvent operations, and create new value. From agent-driven automation to sustainability gains, six PwC leaders share their predictions on what’s coming next.
++
Agentic AI gets real – and benchmarks finally arrive
Dan Priest, Chief AI Officer, PwC US
Agentic AI has gained significant traction over the past year, and 2026 is shaping up to be the moment when its real value becomes clear. Organizations are moving past isolated pilots and driving toward measurable business impact, solving the technical and workflow problems that support achieving their goals.
A measure of maturity and progress is the emergence of AI benchmarks. These are areas of AI investments that can be measured and compared to sector and cross-industry performance. Investment levels, investment areas, and value realization measures are emerging in ways that make comparables possible. It’s time to measure and produce ROIs on many of these AI investments. That will happen in 2026.
Responsible AI evolves from principle to practice
Rohan Sen, Principal for Cyber, Data & Tech Risk, PwC US
Responsible AI (RAI) moves firmly into the “how” column this year. As agentic workflows grow fast – often handling half of what people do today – governance has to keep up. Expect to see more tech-enabled governance from inventory management, cybersecurity, automated red teaming, deepfake detection, and observability built directly into AI systems.
Companies will strengthen alignment across IT, risk, and AI teams, create clearer expectations for human oversight, and modernize documentation and training. When done right, RAI becomes a catalyst for innovation, helping clients achieve their AI ambitions with velocity and trust.
AI becomes a hidden engine for sustainability
Sammy Lakshmanan, Sustainability Principal, PwC US
AI is poised to become one of the most powerful tools for sustainability. While AI usage is growing fast, the way companies use it can dramatically reduce impact. Smart controls, carbon-aware scheduling, and approving compute only when it drives meaningful value can keep energy use in check.
At the same time, AI can help companies hyper-personalize their products and services to customers who value sustainability. It can also optimize transport and energy consumption, strengthen climate resilience, and improve product traceability across supply chains. When sustainability is built directly into AI workflows, both environmental and financial returns grow.
A generalist-driven workforce takes shape
Shebani Patel, Workforce Leader, PwC US
AI is accelerating demand for adaptable generalists – people who can work across the breadth of their function, oversee agents, and turn AI insights into action. They have a wide enough view of the full workflow to guide agents effectively and amplify the value of AI.
Entry-level employees will be prized for their AI fluency, while senior talent focuses on strategy and innovation. As agents take on more routine work, the real shift is toward adaptable generalists who can oversee, guide, and integrate AI across tasks – making sure human insight stays central as workflows evolve. To keep up, organizations will need new skills and incentives that support this shift.
Orchestration layers turn �vibe coding’ into production-ready AI
Jacob Wilson, GenAI Transformation Leader, PwC US
AI experimentation is happening everywhere-and organizations are generating more promising prototypes than ever. But turning these “vibe-coded” ideas into resilient, production-ready systems requires real engineering discipline. Agent orchestration, security, and observability layers will become essential components of every enterprise AI platform.
These orchestration environments function as AI command centers: unifying real-time data across systems; managing drag-and-drop, multi-agent workflows; and integrating multi-vendor LLMs, custom tools, third-party APIs, MCP servers, and agent-to-agent (A2A) communication. They provide continuous monitoring, traceability, and policy enforcement so organizations can operate AI agents safely and at scale.
As these environments mature, both IT teams and business functions will adopt AI-native roles-designing workflows, supervising autonomous agents, and using agentic AI to extend their own capacity. With a robust orchestration layer and a centralized command center, enterprises can confidently scale AI, reduce operational and compliance risks, and convert experimentation into measurable, aligned business value.
Agentic AI becomes the new Managed Services multiplier
Tim Canonico, Managed Services Leader, PwC US
2026 is the year agentic AI stops being a pilot project and starts integrating into the core engine of Managed Services performance. Instead of piecemeal automation, organizations will build structured models that define which activities agents own end-to-end, where humans provide oversight, and how the two continuously hand off work. The winners in the MS arena will be the ones that integrate and orchestrate agents into daily operations-treating them as teammates that amplify human judgment rather than as side experiments.
This year becomes a clear inflection point: providers that get this right will deliver materially better outcomes while freeing their workforce to focus on higher-value scenarios-not just “doing the work,” but improving how the work gets done. Those that don’t will find their margins and relevance under pressure as clients increasingly expect an agent-enabled experience from day one.
To summarize…
2026 is the year AI moves from “trying it” to truly scaling it. PwC leaders expect clear benchmarks for agentic AI, Responsible AI embedded into daily operations, sustainability-driven value creation, a reshaped generalist-plus-strategist workforce, and orchestration layers that turn experimentation into production.
The message for businesses is simple: experimenting isn’t enough anymore. To win with AI, companies must govern it, measure it, engineer it, and align every deployment with their highest-value priorities.
##
