In conversations with customers across industries, trust is the barrier to AI adoption that comes up more often. Leaders and organizations need confidence that AI systems will perform reliably and remain accountable for suggested business actions, protecting sensitive data and treating people fairly. Ethical AI deployment closes that gap. When a system is technically robust, legally aligned, and ethically governed, customers, partners, and employees can scale it with confidence rather than caution.

SAP has updated the SAP Global AI Ethics policy to version 3.0, reflecting the fast-changing AI landscape and lessons drawn from deep-dive interviews with more than 50 international experts. The revision reflects concrete changes made to strengthen the ethical safeguards, accountability, and safety measures within enterprise AI adoption, addressing the substance of business leaders’ concerns, not just the perception of them.

Our approach to AI rests on three words: relevant, reliable, and responsible. For SAP, ethical deployment isn’t a regulatory checkbox or a marketing layer; it’s the foundation that makes the Autonomous Enterprise possible, where intelligent agents coordinate and execute complex workflows while humans remain firmly in control.

Three disciplines for responsible AI

Ethics sit at the very center of SAP’s broader social sustainability agenda, where responsible AI depends on three disciplines working together, not in isolation:

1. Ethics define our values. We set clear moral boundaries for how technology should serve humanity.
2. Security protects our systems and data. We safeguard core enterprise assets against emerging cyber threats and vulnerabilities.
3. Compliance ensures legal alignment. We are proactively preparing for global regulatory standards like the EU AI Act.

Together, they help teams assess whether an AI use case should proceed, what safeguards it requires, and who is accountable for its performance after deployment, ensuring that as business processes become automated, human dignity, fairness, and agency remain intact.

Ethics underpin the Autonomous Enterprise

As AI shifts from assisting people to acting on their behalf, the governance question becomes more urgent. The Autonomous Enterprise represents a repositioning where AI agents no longer just make recommendations, but execute business processes end-to-end, across finance, supply chain, procurement, HR, and customer experience. As CEO Christian Klein put it, “For the mission-critical processes of our customers, ‘almost right’ just isn’t good enough.”

That’s precisely why ethics and governance can’t be an afterthought bolted onto agentic AI. They have to be the foundation it runs on. Every action an SAP agent takes is logged and traceable, so organizations can always know what an agent did, why it did it, and what data it used. Autonomy without that foundation is just an experiment; autonomy built on it is an operating model organizations can trust at scale.

Translating principles into practice

The SAP Global AI Ethics policy, grounded in UNESCO’s Recommendation on the Ethics of Artificial Intelligence, governs the development, deployment, use, and sale of every AI system across the company. SAP’s updated AI Ethics Handbook translates these principles into day-to-day practice. It helps give engineering and product teams a one-stop reference for applying the policy’s guiding principles, from fairness and non-discrimination to human oversight and sustainability.

To bridge the gap between high-level principles and day-to-day engineering, we embed strict operational checkpoints into our product life cycle:

• Mandatory AI ethics impact assessments: Every SAP AI use case must undergo a rigorous evaluation before reaching deployment, based on 10 AI ethics principles. Use cases receive a risk score and the SAP’s AI Ethics Office helps product teams enforce mitigation measures and safeguards in place. Depending on the risks involved, some use cases go through several review steps and may be escalated to the SAP Global AI Ethics steering committee for a final decision.
• Human oversight by design: SAP systems are built to ensure humans remain in control of critical decisions.
• Internal accountability: Compliance is non-negotiable across SAP. All employees acknowledge the policy, and the rules apply universally, including to internal, self-developed AI applications built using tools like Hyperspace AI or Claude Code.

The process provides clear guardrails that are visible to SAP’s stakeholders, including customers and employees.

Leading the industry in AI governance

The demand for transparent governance is urgent, and independent evaluations highlight significant gaps in corporate readiness worldwide. In benchmark research by the Thomson Reuters Foundation and UNESCO assessing 2,972 companies, SAP was among the only 12.4% found to maintain an explicit policy ensuring human oversight in AI systems.

SAP’s leadership in ethical AI has earned external recognition: the company was named a leading positive example in the Collective Impact Coalition for Ethical AI Progress Report. SAP is also an active participant in the World Economic Forum’s Global Coalition for AI and Social Innovation, convened by the Global Alliance for Social Entrepreneurship in partnership with the Schwab Foundation. Working alongside leading technology companies and social innovators, SAP is helping shape responsible AI deployment to accelerate social impact and extend the benefits of enterprise AI to organizations tackling the world’s most complex challenges.

The broader landscape shows how much ground the industry still has to cover. Data from the S&P Global Corporate Sustainability Assessment (CSA) shows that only 36% of responding companies have a dedicated or integrated AI policy. Research from BSI paints a similar picture: just 27% of global businesses currently factor their AI footprint into their broader sustainability strategies, a notable gap given the surging energy demands of AI data centers.

Unlocking sustainable growth

Ethical AI goes beyond risk mitigation and regulatory readiness—it drives tangible economic value. By ensuring fairness, transparency, and robust data privacy, organizations protect their brand reputation while empowering their workforce to perform at their highest level.

When AI coordinates, executes, and optimizes routine business processes within an ethical context, employees are freed from manual tasks to focus on strategy, creativity, and human connection. This synergy between human potential and artificial intelligence forms the backbone of sustainable enterprise growth.

Organizations that understand this will reap the rewards as regulations tighten and societal expectations increase. However, ethical AI only becomes a reality when we work together. We look forward to continuing to work alongside our customers and partners as we embrace one of the most important challenges and competitive advantages in the AI era.

Read the SAP Global AI Ethics policy version 3.0. Find out more about the Autonomous Enterprise.

Matthias Medert is head of Sustainability at SAP.

Swarovski has followed its cloud transformation from 2023 with a global go-live of SAP Cloud ERP Private after choosing an exciting brownfield approach for the rollout. With its migration, the luxury brand is laying the foundation for using AI and for reaching its strategic targets by 2030.

Anyone who is looking for a prime example of how migration to the cloud can do far more than just simplification and standardization should take a closer look at Swarovski. The legendary manufacturer of precision-cut crystals, jewelry, and watches—with its origins in Wattens, in the Austrian region of Tyrol—has transformed its IT landscape from a cost factor into a strategic tool for a digital future.

The transformation was guided by Lea Sonderegger, serving in a dual role as CDO and CIO at Swarovski, with such great success that she was awarded the special “Cloud Excellence” prize in the large enterprise category at the CIO of the Year ceremony held by CIO Magazine last October.

The judging panel found her brownfield approach to be especially praiseworthy: Swarovski employees use SAP Cloud ERP Private but continue to use the familiar processes and databases. A complete redesign of these processes in parallel to the migration would have been too risky and cost-intensive. It would have also resulted in a much longer project duration, Sonderegger is convinced.

25,000 tests with 600 participants

The brownfield implementation was carefully executed. Preparations took two years and involved more than 600 participants performing around 25,000 tests. Two dress rehearsals with strict governance ensured that every function and every data point was ready for the migration.

Sonderegger and her colleagues reserved a 66-hour conversion window for the go-live on April 20, 2026. During this period, all global IT processes at Swarovski were paused. During the subsequent sensitive hypercare phase, 24×7 support ensured that any issues that arose could be dealt with quickly. Thanks to these measures, the transition was seamless. After the conversion window closed, all processes resumed without problems. 

Simplification and standardization ensure consistent data

Despite the large effort involved, this migration was merely the first step. While the switch to SAP Cloud ERP Private created the technical foundation, it’s the subsequent investments that deliver additional added value. These investments concentrate on the incremental reduction of complexity through consolidation of fragmented solutions, the reassessment of user-specific code, and the harmonization of data—all with the overall goal of creating a more coherent, easier-to-handle ERP landscape.

To achieve this, Sonderegger and her team are replacing user-specific applications with SAP standard solutions step by step and only leaving custom developments in place where they offer clear advantages. “The combination of simplification and a return-to-standard solutions improves data consistency, provides for robust, reliable processes, and, ultimately, makes our entire organization more agile,” Sonderegger says.

Cloud technology is not an end in itself

By integrating key functions such as finance, supply chain management, retail, and e-commerce—and enabling their combined use in the cloud—SAP Cloud ERP Private provides for reliable processes and consistent data quality all while enabling customer experiences on a wide variety of front-end solutions on this side of the ERP system.

SAP Cloud ERP Private manages a diverse product range at Swarovski across different regions and price points and integrates with the planning results provided by other SAP and non-SAP systems.

“In all of these activities, cloud technology is never an end in itself, but rather a lever for improving efficiency, resilience, and innovative capabilities,” Sonderegger says. This determination is especially important to her.

It’s not an IT project, it’s a business transformation

Ultimately, Sonderegger and her team succeeded in executing the project on time and on budget because its scope was clearly defined, and strict discipline in change management prevented mission creep. In addition, the company benefited from the experience of its implementation partner, SAP Consulting, and its unrestricted access to SAP expertise.

The example of Swarovski proves that even an essential, unavoidable migration can and should do much more than just avoid risks and cut maintenance costs. The implementation of SAP Cloud ERP Private was imperative here, because SAP ERP Central Component (SAP ECC) had reached the end of its lifecycle.

And the implementation is showing the luxury goods manufacturer the way to the future because everyone involved in the process didn’t just consider it to be an IT project but, above all, a business transformation from day one. One that involved hundreds of experts from different fields and that enjoyed full management support from the beginning.

AI-driven demand forecasts optimize warehouse stocks

Artificial intelligence is also playing a key role in this implementation, with SAP Cloud ERP Private as the operational backbone of an AI ecosystem that can deliver reliable, real-time data and robust, standardized transaction processes.

Swarovski doesn’t use artificial intelligence as a standalone technology, but instead as an integrated capability that complements business processes across all functions. The company is already using AI for demand forecasting, for example, and then uses the results to optimize warehouse stock levels across regions, with the aim of improving the customer experience.

And the AI agent factory initiative enables the development of AI agents that link SAP Cloud ERP Private data with data from non-SAP systems, always with the objective of “automating repetitive tasks, supporting decision-making, and boosting productivity along the entire value chain,” Sonderegger emphasizes.

Top image courtesy of Swarovski

Thirty years ago, SAP launched its certification program to help professionals prove expertise and advance their careers.

At SAP Sapphire, that mission is being redefined for a fundamentally different environment, one in which every industry faces the same core challenge: success depends not just on what professionals know, but on how effectively they can apply that knowledge alongside AI.

Technology has already changed. What now differentiates organizations is not access to innovation, but the ability to translate it into outcomes. According to the World Economic Forum, skills gaps are the primary barrier to transformation, ranking ahead of investment constraints and regulatory complexity. Closing that gap requires more than expanding training catalogs. It requires rethinking how skills are built, validated, and continuously developed.

Certification reimagined

SAP Certification has been redesigned to reflect how work actually gets done. Across more than 100 certifications, traditional multiple-choice exams have been replaced with scenario-based and system-based assessments. Candidates work through case-based challenges, role simulations, and practical tasks in SAP environments that mirror real-world complexity. They can also use AI tools during exams—by design, not exception.

This marks a fundamental shift. Certification is no longer a test of knowledge recall; it is a demonstration of applied capability: the ability to navigate ambiguity, make decisions, and use AI as a tool without relying on it. More than 100,000 exams have already been completed under this model, establishing a new benchmark for certification at scale and reinforcing the relevance of certification in an AI-driven workplace.

Learning is evolving in parallel

In SAP Learning Hub, AI is transforming how professionals engage with content. These capabilities are enabled by the integration of selected functionalities from Google NotebookLM into the customer and partner editions of the platform.

This shifts learning from passive consumption to active interaction. Learners can engage with SAP content in more than 80 languages, ask questions, and receive precise, source-based answers with direct references to official materials. AI also generates complementary formats. Podcasts are available for moments when a screen is neither available nor practical, whether commuting, traveling, or simply stepping away from the desk, available both for passive listening and as interactive conversations with AI hosts. FAQs, study guides, mind maps, timelines, briefing documents, and video overviews allow learning to adapt to individual needs and time constraints.

Early adoption underscores the impact. More than 7,500 users are already leveraging these capabilities. LTIMindtree reports onboarding that is 50 percent faster, while NTT Data Business Solutions has made SAP Learning Hub its primary environment for developing talent prepared for the agentic AI era. The shift is clear: Learning is becoming embedded in daily work, not separated from it.

Building the data foundation

At the same time, SAP is addressing a prerequisite for effective AI: data. Many organizations continue to operate with fragmented and inconsistent data landscapes, limiting the impact of AI initiatives. The Becoming a Certified SAP Data Architect learning journey focuses on building the capability to connect, govern, and structure enterprise data, ensuring that AI systems operate on a reliable and consistent foundation.

This capability is increasingly strategic. Organizations that establish a strong data foundation can move faster from insight to action, scale AI more effectively, and create more consistent business outcomes. In this sense, data architecture is no longer a back-end concern; it is a core enabler of enterprise transformation.

Skills at scale

SAP has committed to equipping 12 million people with AI-ready skills by 2030. Delivering on this ambition requires expanding access while maintaining depth and relevance. Select AI live learning sessions such as All Things Business AI: Latest Insights, Updates & Live Demos, are now available without login or cost, giving professionals at all career stages direct access to SAP’s business AI strategy.

Role-based learning journeys provide targeted development for key profiles such as enterprise architects, while a dedicated “Clean Core” course supports organizations in maintaining SAP S/4HANA landscapes in ways that enable faster innovation cycles and more efficient adoption of new capabilities.

Scaling skills also requires ecosystem reach. SAP’s partnership with Accenture LearnVantage expands global access to instructor-led training, combining SAP-authored content and training systems with Accenture LearnVantage’s proven experience in technology skills development for enterprise clients. This creates a continuous path from foundational knowledge to hands-on experience to certification, reflecting how professionals actually develop skills: progressively, in context, and in alignment with real-world application.

A broader shift

These developments point to a broader shift. Learning is no longer episodic; it is continuous, adaptive, and embedded in how work gets done. Participation in SAP learning has increased by 33% year over year, reinforcing that organizations increasingly view skills as strategic assets in an AI-powered economy. The Autonomous Enterprise takes shape differently across industries, and so does the capability required to make it work.

At SAP Sapphire, SAP marks 30 years of certification not by looking back, but by redefining its role. Certification is becoming a measure of capability in action. Learning is becoming an ongoing process that evolves alongside technology and business needs.

In an AI-driven world, advantage will not come from access to technology alone, but from the ability to apply it with purpose. Across industries, the pattern is consistent: how quickly organizations capture value from AI depends on the people deploying it.

To explore these innovations in more detail and understand how SAP is enabling organizations to build AI-ready skills at scale, read the SAP Sapphire Innovation News Guide.

Andre Bechtold is president of SAP Industries and Experiences.