2025 AI Business Predictions

1. Your AI strategy will put you ahead — or make it hard to ever catch up

AI strategy is about value that starts right now — and this value is not just productivity or efficiency. Some AI systems can now reason independently and “understand” the impact of their decisions. That helps AI perform complex tasks such as designing new services or go-to-market strategies. It also helps AI catch its own mistakes. With AI increasingly powerful and reliable, it's time to embed it in your operational fabric. If you don’t, your competitors who do may establish lasting advantages.

An effective AI strategy, designed to deliver value at scale this year, takes a portfolio approach. One part of the portfolio develops a strong “ground game” to deliver many small wins. It’s a systematic approach that harvests additional value from a growing number of more engaging experiences, higher revenue-generating products and services and more productive workflows. This approach depends on scale, but it also requires carefully setting priorities in a phased approach, with each phase generating value that helps pay for the next. The second part of the portfolio picks some “roofshots,” projects that are attainable but require dedicated attention and resources such as all-new ways of working, interacting with customers or designing products. The third part of the portfolio approach focuses on a few high-reward and highly challenging “moonshots” such as new AI-driven business models. Since the roofshots and moonshots require serious resources — including AI specialists’ time — business owners or the C-suite should choose and lead them.

What won’t matter as much for AI strategy is your choice of large language model (LLM). There will be many good options. Everyone will be using them. A shrewd strategy will instead emphasize what can set you apart — how you leverage AI with your institutional knowledge and proprietary data, with the help of AI-powered cloud architectures.

2. Your workforce could double — thanks to AI agents

If you think AI will shrink your workforce, think again. You’re going to welcome a host of new members to the team this year: digital workers known as AI agents. They could easily double your knowledge workforce and those in roles like sales and field support, transforming your speed to market, customer interactions, product design and so on. An AI agent can autonomously perform many tasks, such as handling routine customer inquiries, producing “first drafts” of software code or turning human-provided design ideas into prototypes. Workflows will fundamentally change, but humans will still be instrumental since game-changing value comes from a human-led, tech-powered approach. People instruct and oversee AI agents as they automate simpler tasks. People iterate with agents on more complex challenges, such as innovation and design. And people “orchestrate” teams of agents, assigning tasks and then improving and stitching together the results.

Thinking about agentic workflow as a fundamental part of your workforce strategy may be a big leap for many companies. It will, for example, involve new management roles responsible for integrating digital workers into workforce strategies, then monitoring and governing them. But the sooner you begin thinking this way — and transforming your operating model to plan, train and manage a blended digital and human workforce — the better positioned you are to capitalize on AI. When you have both digital and human workers on the job, for instance, you can plan for greater agility and shift resources more quickly to meet changing demands.

As AI agents rise, they’ll do in-house some of what you currently outsource. Their advantages go beyond cost savings. You will have greater control, greater ability to customize and a greater ability to please end users. For customer service, AI agents may enable you to offer customers both faster, more satisfying self-service, and better equipped human specialists for high-touch, high-value interactions. AI agents can push the right information in front of your people at the right time so they can quickly and effectively address even complex customer needs. With AI agents, long-term plans for your geographic footprint may need an update. At the very least, consider how your current growth curve for outsourced services will change.

3. ROI for AI depends on Responsible AI

Risk management and Responsible AI practices have been top of mind for executives, as we predicted last year when we said 2024 would be a moment of truth for trust in AI. Yet there has been limited meaningful action. That will change. In 2025, company leaders will no longer have the luxury of addressing AI governance inconsistently or in pockets of the business. As AI becomes intrinsic to operations and market offerings, companies will need systematic, transparent approaches to confirming sustained value from their AI investments. They’ll also need to manage the risks of large-scale deployment. Rigorous assessment and validation of AI risk management practices and controls will become nonnegotiable. Even if the specifics of AI assessment and validation are not mandated, stakeholders will demand it — just as they demand confidence in other decision-critical information (such as financial results) or in cybersecurity or privacy practices.

Business leaders, especially those driving AI transformations, will begin to champion this necessary oversight. They won’t wait for regulatory clarity. AI is moving too quickly and is too business-critical for that. When AI was only in isolated use cases, there was a limit to the damage that disappointing ROI, inaccurate outputs or compliance failures could cause. Now, employees rely on it daily. Customers regularly engage with AI-powered experiences and services. And it will be essential for revenue growth. If AI isn't trusted by stakeholders, if it’s subject to a cyber breach or other risk issue or if initiatives run behind schedule or over budget, your company will take a hit.

To implement AI oversight that unlocks value, you’ll need a second set of eyes. This could come from appropriately upskilled internal audit teams or a third-party specialist conducting an assessment based on leading industry practices and standards. Regardless of how it is achieved, an independent perspective on your AI governance and controls will be critical in 2025 and beyond.

4. AI will be a value play — and a boon for sustainability

AI will accelerate the energy transition. It will also help companies meet their sustainability goals – especially those in emissions-intensive sectors like manufacturing, construction and transportation – if they take the right approach. AI requires so much energy that there’s not enough electricity (or computational power) for every company to deploy AI at scale. More chips are coming, models are advancing and the energy supply is expanding. But we won't hit an equilibrium of supply and demand in 2025. That will make it wise to treat AI as a value play, not a volume one. Use it in more and more areas, yes, but also be strategic about how and where you roll AI out. You can, for example, design AI interfaces to encourage users not to waste AI time and tokens.

But these near-term challenges shouldn’t overshadow the big picture. AI will be a driver for sustainability. Globally, it will likely speed up the shift to renewables. In the United States, neither economics nor stakeholder pressure will permit a massive rollout of new fossil fuel plants. Instead, business will encourage more renewable supplies (including nuclear) and a more modern grid that uses energy more efficiently and brings it where it’s needed. Some pressure may come from your company. Even if AI vendors bear most of AI’s carbon footprint, you are the end user, so it should show up on your carbon balance sheet. To reduce the impact, you’ll want AI vendors to be green.

Inside your company, AI can potentially simplify compliance with a new wave of sustainability disclosure regulations in the United States, the European Union and elsewhere. The November election means the SEC’s climate-related disclosure rules will likely remain on hold, which may create a void that is filled by states following California’s lead and developing their own rules and requirements.

AI can automate internal and external data collection needed to meet these regulations, analyze the data and generate reports (which can be refined by the finance function). AI’s capacity for data collection and analysis will also help you optimize sustainability across your supply chain. Thanks to AI, even small suppliers will be able provide granular sustainability data such as their monthly or annual energy consumption. AI can quantify new kinds of value like the benefits of commercializing low-carbon products. As these AI capabilities are embedded into corporate strategy and everyday enterprise applications, everyone, not just ESG specialists, will be able to access and use sustainability data to help make decisions.

5. AI will cut product development lifecycles in half

If your company makes tangible goods and your product development teams aren’t using AI for design, prototyping and testing, now is the time to start. Multimodal AI — capable of processing and generating diverse data types, from CAD files to simulations — is now revolutionizing product design and broader R&D processes. For example, GenAI tools can propose improved configurations for a car chassis, simulate performance under different conditions and even suggest designs that engineers might have overlooked.

AI can help you iterate designs in hours not weeks, test solutions virtually before building prototypes and troubleshoot more problems before you move to production. Based on PwC’s work with clients and our analysis of technology and industry trends, we’re confident that adopting AI in R&D can reduce time-to-market 50% and lower costs 30% in industries like automotive and aerospace. In many pharmaceutical companies, AI has already helped reduce drug discovery timelines by over 50%.

Most companies are unprepared for this revolution in physical product design. AI is ready to deliver — but the skills gap is often a hurdle. Engineers with deep expertise in design and manufacturing often lack even foundational data science skills. Upskilling these teams and recruiting AI-savvy talent must begin now. Those who embrace AI’s potential in product development will enjoy faster speed to market, lower costs and increased personalization — and that can add up to more satisfied end users.