Digital transformation 2025: What’s in, what’s out

Some argue gen AI’s emergence has rendered “digital transformation” passé. AI transformation is the term for them. Others suggest everything should be called “business transformation” — or just “transformation” for short.

What terminology should you use? The one that drives the greatest call to action from your board, executives, and employees — because maintaining the status quo is a sure path to disruption.

Organizations will always be transforming, whether driven by growth opportunities, a pandemic forcing remote work, a recession prioritizing automation efficiencies, and now how agentic AI is transforming the future of work.

Regardless of the driver of transformation, your company’s culture, leadership, and operating practices must continuously improve to meet the demands of a globally competitive, faster-paced, and technology-enabled world with increasing security and other operational risks.

What’s driving digital transformation priorities in 2025

Digital transformation is not about deploying the latest technologies, addressing technical debt, or advancing infrastructure agility — it’s about how such strategic initiatives drive growth, improve customer experiences, scale workflows, increase quality, and deliver other vital business outcomes. Transformational CIOs continuously invest in their operating model by developing product management, design thinking, agile, DevOps, change management, and data-driven practices.

But this year three changes are likely to drive CIOs’ operating model transformations and digital strategies:

• In 2024, enterprise SaaS embedded AI agents to drive workflow evolutions, and leading-edge organizations began developing their own AI agents.
• The US President-elect promises many changes impacting enterprises, including import tariffs, immigration deportations, energy policy changes, and relaxation of other business regulations that will impact supply chains, labor pools, and other global consequences.
• Driving a curious, collaborative, and experimental culture is important to driving change management programs, but there’s evidence of a backlash as DEI initiatives have been under attack, and several large enterprises ended remote work over the past two years.

Based on those and other criteria, here are three digital transformation practices CIOs might want to increase their focus on in 2025, and three worth replacing with other strategies or practices.

In: Developing transformational leadership and AI-ready employees

One of my three key digital transformation priorities for CIOs in 2024 was developing transformational leadership to help increase the amount of strategic initiatives, experimentation, and change management programs IT can support.

In 2025, AI will continue driving productivity improvements in coding, content generation, and workflow orchestration, impacting the staffing and skill levels required on agile innovation teams. But CIOs will need to increase the business acumen of their digital transformation leaders to ensure the right initiatives get priority, vision statements align with business objectives, and teams validate AI model accuracy. CIOs must also drive knowledge management, training, and change management programs to help employees adapt to AI-enabled workflows. 

“With the rise of AI agents and excitement among the C-suite to stay ahead of new tech developments, IT leaders will face increased pressure and workloads — and democratizing access to AI and upskilling employees will become a bigger priority than ever,” says Ed Macosky, chief product and technology officer at Boomi. “In 2025, businesses intentional with upskilling will maximize AI benefits with a competitive edge, while those who rush to incorporate AI’s next big thing before their team is ready will be hindered in their efforts to innovate.”

Out: Sponsoring moonshot AI innovations lacking business drivers

How much patience will boards and executives have with ongoing AI experimentation and long-term investments? Gartner recently suggested AI is heading for the trough of disillusionment, and two reports imply the AI honeymoon is ending:

• Deloitte’s State of Generative AI in the Enterprise reports that nearly 70% of respondents said their organization had moved 30% or fewer of their gen AI experiments into production.
• Warton’s Navigating Gen AI’s Early Year Report says 57% anticipate slower AI spending increases, “an indicator that enterprises are still searching for ROI on their initial investment.”

“2025 will be the year when generative AI needs to generate value,” says Louis Landry, CTO at Teradata. “That means the gen AI bonanza of investment will slow as companies focus less on simply getting into the AI game and more on real-world opportunities that can generate real value.”

Since many early AI wins drive productivity improvements and efficiencies, CIOs should look for opportunities where real cost savings can drive further innovation and infrastructure investments.

“What’s in is self-funding AI-led business reinvention by slashing technology, data, and process debt and adopting AIOps and AI-enabled software development,” says Amit Bajaj, North America president of TCS. “Realizing business value and tracking capabilities so that the resulting savings can fund innovation, transform the customer experience, reimagine cost structures, and deliver enterprise agility.”

SAS CIO Jay Upchurch says successful CIOs in 2025 will build an integrated IT roadmap that blends generative AI with more mature AI strategies. “As the shine wears thin on generative AI and we transition into finding its best application, it’s more important than ever that CIOs and IT leaders ensure [they are] using AI in a point-specific way that drives business success,” he says.

In: Doubling down on data and AI governance

Getting business leaders to understand, invest in, and collaborate on data governance has historically been challenging for CIOs and chief data officers. Defining policies and other AI governance was a priority at many organizations trying to channel how employees used copilots while protecting sensitive data from leaking to public LLMs. In 2025, CIOs should integrate their data and AI governance efforts, focus on data security to reduce risks, and drive business benefits by improving data quality.

Ravi Ithal, GVP and CTO of Proofpoint DSPM, highlights the importance of a synergistic data and AI governance strategy by thinking of data as the fuel and AI as the engine: “If you’re throwing random fuel types into a high-performance engine, don’t be surprised if it backfires. For AI to deliver safe and reliable results, data teams must classify data properly before feeding it to those hungry LLMs.”

Focusing on classifying data and improving data quality is the offense strategy, as it can lead to improving AI model accuracy and delivering business results. CIOs who struggle to make a business case solely on this driver should also present a defensive strategy and share the AI disasters that hit businesses in 2024 as an investment motivator. 

 “We will see a rise in stories of poor uses of AI, which will cause brands to pump the brakes a bit and revisit their data strategies,” says Bill Bruno, CEO of Celebrus. “Brands struggling to activate AI in meaningful ways because most of their data is unstructured, incomplete, and full of biases due to how digital data has been captured over time on their websites and apps.”

To bring these issues to the forefront of employees’ and managers’ minds, CIOs must explain how dark data and other data debt issues will impact the business competing in the AI era.

“Dark data, often hidden in emails, spreadsheets, outdated systems, and often a derivative of the main data sources, can include sensitive intellectual property or personal data, making it vulnerable to breaches,” says Nishant Doshi, chief product and development officer at Cyberhaven. “AI tools exacerbate the issue by exposing these data pockets, creating new security risks.”

Out: Lift and shift, app migrations, and dumb automations

A positive outcome of AI is that more business leaders will recognize the importance of transforming operations rather than transitioning existing ways of getting work done with newer and better technology. Hopefully, this will lead more CIOs away from lifting and shifting workloads to the cloud, modernizing applications without improving experiences, and implementing purely robotic process automations instead of more business transformative approaches.

For example, migrating workloads to the cloud doesn’t always reduce costs and often requires some refactoring to improve scalability. Brian Singer,  CPO and cofounder of Nobl9, says, “In 2025, organizations will leave behind the mindset of moving everything to the cloud, and as companies double down on reigning in costs, they will start to understand that static workloads may be cheaper to run on-premises. Migrating to the cloud without accounting for nuances of cost and complexity is out, as AI workloads add to operational complexity and are a massive contributor to runaway cloud costs.”

App modernizations that address only technology issues should also be on the way out because using AI copilots, upgrading to gen AI test capabilities, and shifting to platform engineering all simplify constructing better applications. The focus will shift to enhancing user experiences, embedding AI capabilities, and iteratively improving business outcomes.

“AI-boosted development will mean that the user experience designers’ vision will finally be implemented as requested instead of compromised due to a higher level of development effort,” says David Brooks, SVP of evangelism at Copado. “ User feedback will be collected and summarized by AI to inform the next round of improvements, completing the virtuous cycle.”

Other coding efficiencies and improvements in developer experience should give CIOs new incentives to drive stronger software development practices and reduce technical debt. This upskilling will be critical as organizations look to leverage proprietary data in LLMs, develop AI agents, and safely integrate capabilities with their ecosystem of partners.

“The future of applications is composable in that APIs are the conduit for AI integration, and AI enables APIs by providing the intelligence to enhance functionality and efficiency of API interfaces,” says Jason Gartner, general manager of product management application development and integration at IBM. “This new generation of composable applications, driven by the remarkable performance of modern development tools and techniques, will help organizations accelerate delivery of new products and solutions to drive growth and improve experiences.”

In: Increased training on security, safety, and trust

If CIOs and CISOs find it challenging to educate employees to recognize malicious emails and train them not to click on links from unknown sources, then the new wave of AI threats will require doubling these efforts.

“In 2025, the proliferation of increasingly sophisticated deepfake technology will begin to pose a significant threat to trust in digital content and online interactions,” says Atticus Tysen, CISO and CIO at Intuit. “As deepfakes become more accessible and harder to detect, individuals and organizations will need to develop new strategies for verifying information and ensuring authenticity in the digital age, including educating and promoting awareness for the workforce.”

Mike Arrowsmith, chief trust officer at NinjaOne, says that as AI evolves, it will get even better at data attribution, making it more difficult for organizations to distinguish between real and malicious personas. “In 2025, there will be an increased and renewed focus on IT and AI employee training that educates staff on identifying AI risks to ensure organizations are prepared to address the security gaps AI proliferates.”

Law Floyd, chief of security operations at Telos, adds this third dimension when considering security, safety, and training programs. “Insider threats will continue to be an issue for organizations, and one of the best ways to combat them is through an effective training program.”

Out: Expecting employees will keep up with AI

The fundamental mistake in digital transformation is underinvesting in change management or waiting until the eleventh hour to plan for it. At the recent Spark Executive Forum, I led a panel that discussed how accepting a business-as-usual mindset is a barrier for CIOs to address because of AI’s opportunities to impact productivity. AI capabilities are accelerating the pace of change, leaving many employees behind in adopting new technologies and adapting to AI-enabled workflows.

The pace of change requires organizations to support lifelong learning and go beyond skill-based training to a culture that supports experimentation, teaching, and delivering continuous improvements. CIOs will need to evolve their cultures significantly in 2025 as the impacts of gen AI, political changes, regulations, and culture challenges can either disrupt or accelerate a business’s ability to compete.