AI is moving fast – faster than most organisations can adapt.
While SAP is rolling out hundreds of new AI capabilities across its portfolio, many companies are still struggling to turn AI ambition into real business value. Adoption, data quality and compliance often become bigger obstacles than the technology itself.
To better understand what’s actually happening on the ground, we interviewed Michael Wolff, Chief AI Officer at Pearl Group, about how SAP customers should think about AI today and what companies should focus on if they want results.
We interviewed Pearl’s Chief AI Officer
Michael Wolff is Chief AI Officer at Pearl Group with 26+ years of AI experience and 20+ years with SAP, including 13 years at SAP themselves. His career spans AI product strategy, M&A, and enterprise architecture across Europe, Asia and Silicon Valley.
At Pearl, his role is to help customers close the gap between powerful SAP technology and real-world adoption.
“There is no shortage of AI technology today. The real challenge is making it work inside organisations – in a compliant, secure, and value-driven way.”
AI is just a tool - but it needs boundaries
AI is often described as something almost human-like. Michael prefers a more grounded perspective.
“AI is a tool. Powerful, but immature. Without guardrails, it behaves like a five-year-old with access to your most sensitive data and business-critical decisions. That's not a criticism of the technology. That's reality. The organisations that treat AI as magic will fail. The ones that treat it as infrastructure, with boundaries, governance, and human oversight, will win.”
That immaturity is exactly why guardrails are important, especially in enterprise systems.
“AI doesn’t understand context or consequences the way humans do. That’s why governance, controls and clear boundaries are absolutely critical – particularly in business-critical systems like SAP.”
Technology is ahead of organisations
One of the most striking parts of the conversation is the gap between AI investment and actual business impact.
“Last year alone, companies globally spent around 1.5 trillion dollars on AI. 88% of companies use AI in at least one function. Only 39% attribute any EBIT impact. Only 6% achieve significant enterprise value. 42% abandon AI initiatives before production. That number was 17% last year. 95% of generative AI pilots remain stuck in pilot purgatory.”
According to Michael, this is not a technology problem.
“The technology is already there. SAP is delivering AI at scale. What’s lagging behind is organisational readiness – processes, data structures, skills, and decision-making.”
Michael Wolff, Chief AI Officer, Pearl Group
Consumer AI is setting enterprise expectations
"Two weeks ago, an open-source AI agent called Moltbot hit 145,000 GitHub stars. Best Buy sold out of Mac minis because people wanted dedicated hardware to run personal AI assistants around the clock. The tagline was 'AI that actually does things.'
That expectation now flows into the enterprise. Employees wonder why approving an invoice takes 47 clicks when their personal AI books flights and manages calendars autonomously. SAP has responded with 30 Joule Agents across finance, procurement, and operations, but the pressure from consumer AI is real and accelerating."
Where AI delivers value in SAP today
Despite the challenges, AI is already creating real value for SAP customers – particularly in finance.
“Finance is where we see the strongest and fastest impact today. Cash management task time drops by 70%. Master data management sees 85% effort reduction. These are SAP's documented benchmarks, and we see them replicated across Pearl's customer base.”
Other areas are following quickly, especially where large data volumes and repetitive tasks are involved.
“The biggest wins come when AI supports decision-making and removes tedious work – not when it tries to replace humans entirely.”
Adoption – not fear – is the real barrier
Public discussions about AI often focus on fear: job losses, complexity, or loss of control. Michael sees something else holding companies back.
“In my experience, fear is rarely the real problem. The real issue is adoption.”
He continues:
“Two failure modes explain why. Technology without adoption: features get implemented, nobody uses them. Adoption without technology: people trained on tools that don't integrate with core systems. Both fail. Technology and adoption must happen simultaneously.
Why wouldn’t you want a tool that removes repetitive, boring tasks and frees people up to focus on more meaningful work? The challenge is helping organisations actually use the tools they already have.”
This is where Pearl often steps in.
“We ran our own transformation first. Pearl deployed AI across 500+ consultants and measured 70% productivity gains before offering services to customers. We became case study zero. That credibility matters when we ask customers to trust us with their AI adoption.”
Compliance is the AI risk many overlook
One of the most underestimated aspects of AI, according to Michael, is compliance.
“When you enable certain AI features, your data may no longer be your data. That’s something many organisations don’t fully realise.”
With regulations like GDPR, this becomes critical.
“You need to understand where your data goes, how it’s processed, and under which legal framework. AI doesn’t remove responsibility, it increases it.”
This is an area where SAP’s approach to enterprise AI stands out, but it still requires careful implementation.
“SAP provides strong frameworks for responsible AI, but customers need guidance to use them correctly.”
Don’t expect AI magic without a clean SAP core and solid data
AI success in SAP is closely tied to data quality and system architecture.
“If you don’t have a clean core, AI will never work. It’s as simple as that. Garbage in, garbage out.”
Many organisations underestimate how foundational this is.
“You can’t layer AI on top of heavily customised systems and expect magic. The groundwork has to be done first.”
Closing the SAP AI adoption gap
Looking ahead, Michael believes the organisations that succeed with AI will be those that treat it as a long-term capability rather than a quick experiment.
“The companies that succeed will be the ones that decide to implement AI and adoption at the same time.”
In practice, this means working across technology, data, compliance and people in parallel.
“AI doesn’t fail because the algorithms are bad. It fails because organisations underestimate what it takes to adopt it safely and at scale.”
This is where Pearl’s AI-related offerings come into play. Rather than leaving customers alone once AI functionality is activated in SAP, Pearl supports adoption as an ongoing capability.
One example is Adoption as a Service, where Pearl helps customers operationalise new SAP and AI functionality - from user enablement and process alignment to change management and governance.
“Our customers already have powerful AI features available in SAP, but they’re not being used. Adoption as a Service is about making sure that what’s technically possible also becomes operational reality.”
Another critical area is testing and validation. As AI becomes embedded into business-critical processes, testing can no longer be treated as a one-off project phase.
With UAT as a Service, Pearl supports customers with structured, scalable user acceptance testing - ensuring that new AI-driven processes work as intended, comply with regulations, and are trusted by the business before going live.
“If users don’t trust the outcome, they won’t use it.”
By combining SAP expertise with services focused on adoption and validation, Pearl helps customers move from AI ambition to measurable business value without underestimating the organisational and compliance requirements involved.
“AI in SAP is no longer about the future. It's here. The organisations that implement technology and adoption simultaneously will pull ahead. The ones that wait for AI to mature will find themselves five years behind competitors who started small, measured everything, and scaled what worked.”
