BPM Skills in 2026 (part 2)
Blog: BPM Tips
Part 2 of the post from the series about the BPM Skills is now available.
Check out the thought-provoking answers from 10+ BPM experts.
As always, you can either read everything or use the navigation below. Enjoy!
BJ Biernatowski
Marlon Dumas
Renata Gabryelczyk
Paul Harmon and Vahid Javidroozi
Thomas Hildebrandt
Michael Hill
Martin Holling
Sandeep Johal
Emiel Kelly
Guillermo Lopez
Matúš Mala
Morten Marquard
Which BPM skills will be hot in 2026
Now, let’s dive into the answers.
BJ Biernatowski
BJ Biernatowski is a digital transformation leader specializing in AI-driven process optimization, intelligent automation, and global operations. He has spearheaded large-scale initiatives at Microsoft, Amazon, UnitedHealth Group, and Nordstrom, consistently delivering measurable impact. His expertise spans process modeling, AI-assisted decision-making, and integrating emerging technologies across complex ecosystems.
Passionate about blending strategy with innovation, BJ designs scalable systems that accelerate agility and long-term competitiveness.
WWW: LI profile
[email protected]
How do AI and other trends impact the way organizations manage and run their processes?
AI is accelerating demand for automation, but it’s also exposing gaps in how organizations design and manage their processes. Many teams are dropping AI into the middle of operations without the process architecture, governance, or delivery discipline needed to make it successful. The result is predictable: user pushbacks, inconsistent outcomes, and solutions that don’t scale.
At the same time, the top-down push for “more AI everywhere” often outpaces the operating model needed to guide workers on how to apply these tools responsibly and effectively. Without clear workflows, roles, and guardrails, AI becomes fragmented and difficult to integrate into day-to-day work.
Interestingly, the most successful AI adoption is happening bottom up. Individual practitioners are figuring out how to use AI to extend themselves, close skill gaps, and take on more responsibility. Their success highlights the opportunity and the need for organizations to build the process foundations that allow these wins to scale across the enterprise.
What are the skills, techniques, behaviors, and attitudes that can help BPM practitioners create value for their organizations in 2026?
AI gives experienced BPM practitioners a huge amount of leverage: faster analysis, better pattern recognition, and the ability to step into work that used to require years of experience. But it doesn’t magically give people the skills needed to adopt and scale these technologies. If anything, it makes the gaps more visible. The organizations creating real value in 2026 are the ones pairing AI with strong process fundamentals. Companies trying to “go faster” without redesigning workflows or strengthening their operating models are running into predictable issues: hallucination-driven errors, unclear system behavior, heavier workloads, and the wear and tear that comes from accelerating work without improving it.
For practitioners, the opportunity is massive. AI flattens access to knowledge-intensive parts of BPM and digital transformation, letting people move into areas they haven’t touched before. But the differentiators are still human: process literacy, critical judgment, the ability to design and govern AI-enabled workflows, and the discipline to apply these tools responsibly. Those are the skills that turn AI from a cool tool into a real operational advantage.
What are the best resources to learn those skills? (e.g. books, articles, courses)
The best resources depend on the AI Copilot or automation platform your organization uses. We’re back in a world where vendor ecosystems matter. Microsoft, AWS, UiPath, and others all offer structured academies, hands-on labs, and certifications that map directly to the tools practitioners use every day.
University programs and executive courses can be useful, but they’re expensive and often too theoretical for practitioners who need to design, build, and run AI-enabled workflows. The work itself, infusing AI into workflow automation engines like Microsoft’s Copilot ecosystem, or UiPath’s automation fabric, is technical and requires direct access to the technology. Real learning happens inside the platforms themselves.
For most practitioners, the optimal path is vendor academies backed by certifications and hands-on experimentation. That combination builds practical, platform-specific skills that translate directly into value for the organization.
Which skills are no longer relevant or not practically applicable yet (hype)?
Portions of this answer were developed with the help of AI:
“AI strategy” without operational grounding
High-level AI strategy frameworks that don’t connect to actual systems, data, or workflows sound impressive but rarely lead to implementation. Organizations need practitioners who can execute inside the platforms, not just talk about AI at a conceptual level.Fully autonomous AI agents replacing human oversight
There’s significant hype around “hands-off” AI agents that can independently design, build, and deploy workflows. In practice, no enterprise platform allows this without strict governance, human review, and guardrails. The idea is interesting, but it’s not something organizations can safely operate today.“No-skills-needed” AI development
The narrative that AI eliminates the need for technical, architectural, or process skills is misleading. AI accelerates work, but it doesn’t replace the need for process modeling, data quality, governance, workflow design, or integration fundamentals. The belief that AI can compensate for weak foundations is hype that sets teams up for failure.Prompting and generic AI literacy as career differentiators
Prompting is becoming table stakes, not a specialty. Platforms are rapidly abstracting it behind copilots, templates, and automation patterns. Similarly, standalone “AI fundamentals” courses disconnected from actual platforms have limited practical value – they don’t teach practitioners how to build or deploy anything. The durable skills remain workflow design, data modeling, integration, change management, and governance. Those are the fundamentals that let practitioners execute, not just discuss.
If you want to deliver value to organizations in 2026
Agentic AI is moving into a broader adoption phase in 2026. Definitions vary by platform, but the pattern is consistent: most agents still operate inside a single vendor ecosystem, with early signs of cross-agent integration emerging. That power comes with tradeoffs, especially vendor lock-in, but the capabilities go far beyond what traditional RPA can deliver. Many RPA use cases will be absorbed by agentic AI because agents can reason, adapt, and operate across workflows instead of following rigid scripts.
This shift aligns directly with the three phases of AI-enabled processes we outlined in Practical Business Process Modeling and Analysis (Misiak, Sinur, Biernatowski, 2025):
Phase 1: Smarter resources. Humans, systems, data, and machines are augmented with AI — pattern recognition, generative assistance, and supervised learning. AI accelerates work, improves decision-making, and frees higher-skilled resources by pushing more tasks to augmented workers and systems.
Phase 2: Smarter execution. AI begins to displace time-constrained or high-precision human work with always-on bots and snippets. Humans remain essential where judgment, empathy, and oversight are required. This is a semi-supervised world where processes and people validate AI outputs and maintain control.
Phase 3: Smarter orchestration. AI becomes the decider and controller for processes. Goals and guardrails replace step-by-step instructions. AI dynamically creates process paths, allocates work, and manages bots in real time. Process models shift to an after-the-fact role for transparency, auditability, and explainability.
Agentic AI is essentially the early expression of Phase 3. Vendors have already repositioned their portfolios around this trend, and organizations that want value in 2026 need to prepare for it. That means building living process architectures (Digital Twins), strengthening operating models, and ensuring clear guardrails so agents can operate safely and effectively.
Process models, decision models, and audit trails will remain critical, not as design artifacts alone, but as the transparency layer that explains AI decisions, supports compliance, and helps organizations manage bias, privacy, and emerging legal requirements.
The bottom line: delivering value in 2026 requires understanding where agentic AI replaces legacy automation, how it collaborates with processes and humans, and what governance is needed to keep it aligned with business goals. And as AI investments scale, the ability to demonstrate and measure business value will continue to be one of the most important skills practitioners can bring to the table.
Prof. Marlon Dumas
Marlon Dumas is Professor of Information Systems at University of Tartu and Chief Product Officer at Apromore – a company dedicated to developing process mining and AI-driven process optimization software. While continuing to grow the Apromore product, he conducts a research backed by the European Research Council with the mission of developing AI-based techniques for automated discovery of business process improvement opportunities. He is a widely published researcher, having co-authored over 350 scientific articles, 10 patents, and a textbook (Fundamentals of Business Process Management) used in more than 400 universities worldwide.
WWW: Company website
WWW: LI profile
How do AI and other trends impact the way organizations manage and run their processes?
Automation is going to be again on the spotlight, driven by developments in the field of generative and/or agentic AI. There is going to be a lot of initiatives to automate at two levels. First of all, we will see a lot of automation at the level of tasks, like filling in details for a purchase order or for an invoice. Second, we will start seeing automation happening at the level of end-to-end process orchestration, like triggering API calls to automate the steps in an account opening process in a bank, or automating the orchestration of an invoice handling process.
The difference with respect to previous automation waves is that this time, automation will go beyond the level of inputting structured data. If you think about robotic process automation, it was mostly about entering data into fields in a form or in Excel sheets, by copying data from other fields of spreadsheet cells. This time, automation will also involve unstructured data, such as an AI agent reading from unstructured document and producing structured or unstructured data out of it. We are also going to see automation of certain types of repetitive decisions. These are all capabilities within the purview of agentic process automation.
What are the skills, techniques, behaviors, and attitudes that can help BPM practitioners create value for their organizations in 2026?
On the skills side, the most important class of skills will be and will remain critical thinking skills. The world of AI will be fertile for critical minds, who put into question ideas and thoughts that look right on the surface, but turn out to be misleading or inaccurate once you put a magnifying glass on them.
Tool-relevant specialized skills will also become very important. We are going to see a lot of new types of tools for agentic automation and orchestration coming out. Be ready to analyze their capabilities critically, and to conduct assessments and proof-of-concepts to determine if these tools really address the use-cases you need to implement.
Expertise in specific industry verticals will become highly valuable, such as deep domain expertise in financial processes, field services processes, or logistics processes.
What are the best resources to learn those skills? (e.g. books, articles, courses)
I recommend looking at the references and pointers provided in the manifesto on AI-augmented BPM systems and more recent papers on agentic automation:
- AI-Augmented Business Process Management Systems: A Research Manifesto: https://dl.acm.org/doi/10.1145/3576047
- Agentic Business Process Management Systems: https://arxiv.org/abs/2601.18833
Which skills are no longer relevant or not practically applicable yet (hype)?
Skills in rule-based and script-based automation, such as Robotic Process Automation (RPA), have now become commodity. Skills on GenAI-based or agentic automation, are gaining a lot of traction.
Prof. Renata Gabryelczyk
PhD, DSc, an Associate Professor at the University of Warsaw. She is Head of the Department of Management and Information Technology at the Faculty of Economic Sciences, University of Warsaw. Her academic experience includes involvement in research projects, research fellowships at several universities in Germany and Austria, and numerous publications in national and international publishers. Her research interests include business process management, performance management, facility management, and IT applications. She is a member of the program board of the Polish Certificate of BPMN at the Polish Academy of Sciences, a member of Polish Scientific Society of Economic Informatics, a member of the Technical Committee for Facility Management of the Polish Committee for Standardization, and a member of Polish Chapter of AIS (PLAIS). She serves as Managing Editor in the Central European Economic Journal and as Senior Editor in the Information Systems Management journal.
WWW: LI profile
How do AI and other trends impact the way organizations manage and run their processes?
I agree with the increasingly repeated thesis that the real impact of AI on BPM is currently often overestimated. While technologies such as hyperautomation and agentic AI undoubtedly expand the potential of BPM, many organizations still have not addressed fundamental issues such as a coherent process architecture, the quality of process data, and alignment between BPM objectives and the organization’s strategic goals. In many organizations, advanced technologies are adopted faster than core management capabilities mature. As a result, AI initiatives often reinforce existing weaknesses rather than resolve them. The expected return on investment in AI fails to materialize due to the lack of solid organizational foundations, structured data, and effective governance. Perhaps we should return to the basics and avoid automating chaos.
What are the skills, techniques, behaviors, and attitudes that can help BPM practitioners create value for their organizations in 2026?
Effective BPM will increasingly require the integration of three approaches: process thinking, data-driven thinking, and strategic thinking. Proficiency in working with data, as well as in applying the methods and tools of the full intelligent BPM cycle, is of course essential. For a successful integration of the process perspective with data analytics, communication between data specialists and process experts is key. Such collaboration remains rare in many organizations, limiting the ability to translate BPM competencies into real organizational value.
What are the best resources to learn those skills? (e.g. books, articles, courses)
The most effective learning comes primarily through the practical application of theory. Hands-on experience with real-world processes, data analysis, and the implementation of improvements allows students to understand limitations, trade-offs, organizational culture, and the specific of the business environment. However, access to high-quality educational resources remains limited. It is difficult to design academic courses that prepare students for the realities of organizational complexity and chaos. Moreover, there is a lack of materials tailored to specific industries. BPM in local government, manufacturing, or small businesses requires different approaches and practices. As an academic teacher, I still believe that universities should provide a solid, ideally interdisciplinary foundation. Enabling students to work on projects in real-life conditions offers an excellent springboard for employment. Yet, achieving this continues to rely on close collaboration between academia and business.
Which skills are no longer relevant or not practically applicable yet (hype)?
Highly detailed models that are detached from decision-making or strategic intent tend to become documentation artifacts rather than management instruments. The capabilities of AI are also often overestimated. AI does not understand strategy and cannot take responsibility for decisions. That is why it is essential to take a critical approach to technology and focus on real business and process needs.
Paul Harmon
Paul Harmon is a Co-Founder, Executive Editor, and Senior Market Analyst of the Business Process Trends website – www.bptrends.info – an internationally popular website that provides a variety of free articles, columns and book reviews on trends, directions and best practices in business process management.
In 2003 Paul authored Business Process Change: A Guide for Business Managers and BPM and Six Sigma Professionals (Published by Morgan Kaufman, who issued the fourth edition in 2019).
Paul is also a Co-Founder and a Principal Consultant of BPTrends Associates (BPTA), a professional services company providing executive education, training, and consulting services for organizations that are interested in understanding and implementing business process management.
Paul ’s involvement in business process change dates back to the late 60’s when he worked with Geary Rummler, at Praxis Corp., and was responsible for managing the overall development and delivery of the performance improvement projects undertaken by that company. During the 70s and 80s he ran his own company, Harmon Associates, and undertook major process improvement programs at Bank of America, Security Pacific, Wells Fargo, Prudential, and Citibank, to name a few.
During the same period he was a Senior Consultant at Cutter Consortium and edited their Expert System Strategies, Object-Oriented Strategies, and Business Process Reengineering Strategies newsletters. His book, Expert Systems: AI for Business, coauthored with David King, was a worldwide best seller during the 80-90s. and he consulted with many companies as they explored the uses of Artificial Intelligence during that period.
Paul Harmon is an acknowledged thought leader who is concerned with applying new technologies and methodologies to real-world business problems. He is a speaker and has developed and delivered executive seminars, workshops, briefings and keynote addresses on all aspects of AI and BPM to conferences and at major organizations throughout the world. He is very excited to be following the latest developments in Neural Network-based AI and BPM as they are now being integrated.
WWW: www.bptrends.info
WWW: LI profile
Twitter: @harmon_bptrends
Dr Vahid Javidroozi
Vahid Javidroozi is an Associate Professor in Smart City Systems and Digital Transformation at Birmingham City University (UK), where he is based in the College of Computing, Engineering and the Built Environment. His work focuses on business process management, enterprise systems, digital transformation, and the application of artificial intelligence in complex socio-technical systems.
Vahid’s research spans BPM, ERP systems (including SAP), AI-enabled workflows, large language models, digital twins, and systems thinking, with a strong emphasis on practical impact across sectors such as smart and sustainable cities, supply chains, transport infrastructure, and healthcare. He has published extensively in peer-reviewed journals and conferences, and his work is widely cited in the areas of BPM, smart cities, and AI-enabled digital transformation.
He is the founder and lead of the Smart, Sustainable and Green (SSG) Research Alliance, an interdisciplinary initiative that brings together academia, industry, and public-sector organizations to address urban challenges through systems-oriented, process-driven, and technology-enabled approaches. Through this work, he has led and contributed to numerous UKRI, Innovate UK, Horizon Europe, and international research and enterprise projects, including large-scale collaborations with government bodies, infrastructure operators, and technology partners.
Vahid is an invited member of the BridgeAI Standards Working Group, contributing to national discussions on AI standards, governance, and responsible adoption. He is also a certified Responsible and Ethical AI expert and has been actively involved in translating AI capabilities into organisational processes that are transparent, accountable, and value-driven.
Alongside his research, Vahid has extensive experience in executive education and professional training. He teaches and leads enterprise-focused modules on BPM, ERP, and digital transformation, and works closely with industry partners to support organizational change initiatives. He has supervised and mentored doctoral researchers, early-career academics, and practitioners, with a strong focus on systems thinking, design science research, and real-world impact.
Vahid’s work is driven by a long-standing interest in how processes, people, data, and technology interact within complex systems. He is particularly interested in the evolution of BPM from process improvement within individual organizations toward large-scale, cross-organizational coordination enabled by AI and digital platforms.
WWW: LI profile
How do AI and other trends impact the way organizations manage and run their processes?
We wrote in the 4th edition of Business Process Change and have written several times on the Business Process Trends website that AI is the most profound change that organizations will need to address over the course of the coming decades.
That said, it’s important to note that AI is not a single technique — it’s a large collection of new techniques that can be used in different ways and in various combinations to solve problems. AI systems include human reasoning applications, intelligent robotics, intelligent vision systems, intelligent voice systems, and much more. Like all computer software applications, AI will be integrated with existing business processes to make those processes more efficient and effective.
Think of just one possible application: automated trucks that can move materials from one warehouse to another without a human driver. Such an application would involve specific applications of robotics to load the truck in an efficient manner, an application to see the road and the environment around the truck and to provide information on what’s happening in real time, applications to define the location of the truck (GPS) and to plan its course forward toward some goal, an application to define and enforce laws of the road, an application to quickly define changes in the environment that require changes in plans (an emergency stop, for example), robotic devices to control the steering and movement of the truck and management systems to direct them. It would also require an overall management system to coordinate everything, and perhaps talk with people having questions. Complex visual, robotic and reasoning systems will need to be created and integrated to generate a safe, reliable automated truck that a business will feel confident to use.
In reality, of course, the company managing the use of the warehouse and the trucks will have nothing to do with developing or integrating AI into the driverless truck. They will buy the truck from a vendor and it will come with AI enhancements, just as it comes with a motor or a radio. The warehouse company will need to worry about dealing with transitioning from its existing trucks and drivers to driverless trucks: how to schedule them, maintain them and deal with problems associated with their use. In other words their main concern will be with redesigning the trucking/warehousing process.
In passing, while interested in how AI and process improvement work together, we have also become fascinated in the broader BPM transition between what we term (1) first generation process work — process change that focuses on specific process improvement with a specific business environment (improving or automating an auto production line, for example) and what we increasingly refer to as (2) second generation process work — processes that integrate multiple business processes within or across companies to allow more complex coordination. A worldwide supply chain involving several companies that change in response to real time events provides an example of such a second generation process. While logically independent, we are convinced that AI techniques will increasingly become the key to the design of second generation business processes. Teaching the skills and analytic techniques to facilitate the design and implementation of such second generation processes will be a key challenge to the next generation of process practitioners. And many will require a knowledge of AI techniques to make it possible.
From our perspective, this shift also brings BPM into closer contact with socio-technical complexity. AI does not simply automate tasks; it reshapes decision rights, accountability, and coordination across people, processes, data, and technology. The real BPM challenge is not “adding AI” to existing processes, but redesigning processes so that humans and intelligent systems can collaborate effectively, transparently, and responsibly in increasingly uncertain environments.
What are the skills, techniques, behaviors, and attitudes that can help BPM practitioners create value for their organizations in 2026?
The techniques involved in the development of specific AI applications of all kinds are of little concern to business process practitioners. As far as process professionals are concerned, AI is just a collection of new software and IT techniques that allow them to improve (automate) business processes — just as relational databases, in their time, simply provided a better way to access data and relationships between data. The challenge for process practitioners is to identify opportunities to use AI techniques for process improvement, and then to work with IT to create and implement new systems that incorporate those new improvements.
What are the best resources to learn those skills? (e.g. books, articles, courses)
This is a new area, and there is a lot of nonsense being offered as wisdom. Keep in mind what we have said. Process work is process work. Automating processes using computer applications is something we already know how to do. AI just provides a lot of new automation options. The key is to learn what can be done, today, with the AI techniques currently available. Reading articles and attending conferences — studying case studies — is the best way forward right now.
In particular, practitioners should seek out examples that include both successes and failures, as many AI initiatives fail due to poor process design, unclear ownership, or unrealistic expectations rather than technical limitations.
- Book: Harmon, Paul. Business Process Change (4th ed.). General introduction to process work with a chapter that focuses on using AI in BPM projects.
- The 5th edition, currently in preparation, will expand this, especially in relation to AI-enabled processes and large-scale coordination.
- Javidroozi, V., Tawil, A.-R., Azad, R. M. A., Bishop, B., & Elmitwally, N. S. (2025). AI-Enabled Customised Workflows for Smarter Supply Chain Optimisation: A Feasibility Study. Applied Sciences, 15(17), 9402. https://doi.org/10.3390/app15179402
Which skills are no longer relevant or not practically applicable yet (hype)?
It isn’t a matter of identifying process techniques that are no longer relevant — since all techniques are useful in the context for which they were designed. It’s more a matter of looking at what a given organization is emphasizing today. If you are still working on measuring specific small-scale processes (e.g. processing an order by hand or operating an auto production line with human workers), Lean or Six Sigma may be very relevant.
Most organizations, however, have completed their basic process analysis work — by themselves or by buying applications from companies like SAP. Their emphasis today is on integrating and managing large scale processes — like whole value chains — that stretch across whole organizations, or even multiple organizations to integrate their responses in more-or-less real time. This is an area in which AI techniques are going to prove incredibly valuable.
There are a few organizations that have the people and the knowledge to explore these challenges today. Most do not and to urge them to do so would be to urge them to attempt efforts that would probably end in failure.
For most organizations, this is a time for exploration. Hire new people with some AI experience. Launch small-scale projects that involve AI applications. Grab the low hanging fruit. Attend conferences and listen to what the leading companies are doing. And plan.
Much of the current hype assumes that technical capability automatically implies organizational readiness. In practice, fully autonomous, end-to-end AI-managed processes remain aspirational for most organisations. The near-term value lies in augmentation, learning, and resilience rather than wholesale replacement of human judgement.
Prof. Thomas Hildebrandt
Thomas Hildebrandt has since 2018 been full professor at the Department of Computer Science, Copenhagen University and founder of the research section for Software, Data, People and Society. Thomas has been working as PI and co-PI on inter-disciplinary research and development projects jointly with industry partners in the area of technology and methods for business and workflow management systems for more than 20 years and has and has been a senior PC member of the BPM Conference for several years. Thomas initiated the research on DCR Graphs in 2008 and has since then led the research in collaboration with his research groups and Morten Marquard, the CEO at DCR Solutions. Thomas is also an active speaker on AI and digitalization for industry and public sector organisations and is member of the Danish Standards group for AI, who is part of the European (CEN/CENELEC) and Global (ISO) standardization bodies.
WWW: LI profile
How do AI and other trends impact the way organizations manage and run their processes?
The introduction of widely available LLMs and chatbots based on such has resulted in high expectations from both citizens and directors towards enabling conversational interfaces to the business processes of organizations and companies both internally and externally. While RAG (Retrieval Augmented Generation) solutions dominated the scene last year and still are being tested in many places, the new buzz is agentic AI, where the use of LLMs is no longer limited to question answering but promoted to carry out processes. However, while the introduction of a chatbot is celebrated in the news, many, if not most, are subsequently silently removed because they go off the track.
What are the skills, techniques, behaviors, and attitudes that can help BPM practitioners create value for their organizations in 2026?
A fundamental attitude towards AI for BPM that can help BPM practitioners to create value for their organizations in 2026 is to cut through the hype an realize the fact, that LLMs by design are unrealible and therefore should not be trusted to control any process nor to answer questions that cannot be verified by other means. This does not mean that language models are useless: LMs can be used to generate drafts of business processes from natural language descriptions and also to develop natural language user interfaces to knowledge based, symbolic AI models, rule and process engines, if one ensures a human in the loop to validate respectively the generated process drafts and the translated user inputs. The former is an example of AI for the engineering of business processes, which is most efficient if the target modelling language is close to natural language and has a formal semantics or execution and validation engines (making it possible to automate the validation of the generated processes), such as the declarative DCR Graphs language. The latter is an example of neuro-symbolic AI, or using a less hyped term: Hybrid-AI.
What are the best resources to learn those skills? (e.g. books, articles, courses)
The Hybrid-AI approach is described in https://research.nvidia.com/labs/lpr/slm-agents/. The failures of LLMs for reasoning (and thus trustworthy execution of processes) is described in https://arxiv.org/abs/2602.06176. Information about the DCR graphs technologies can be found here: dcrsolutions.net. The use of DCR graphs for legal reasoning is described in a chapter of the recent book: https://www.routledge.com/Artificial-Intelligence-Humans-and-the-Law/PalmerOlsen-LivingstonSlosser-AddoRavn-Eddebo-HultinRosenberg/p/book/9781032934556 along with other chapters on the use of AI for Law. The use of LLMs for translation of law into symbolic DCR Graph models and then using LLMs to develop a natural language user interface is the goal of the XHAILe research project initiated in 2025: https://di.ku.dk/english/research/research-projects/xhaile/
Which skills are no longer relevant or not practically applicable yet (hype)?
A skill that has never really been relevant for professional use is that of prompt “engineering”, which is a misnomer from the outset, since you cannot engineer something that is not grounded in scientifically validated laws or rules.
Michael Hill
Michael Hill is an experienced editor and journalist. He is the former editor of PEX Network overseeing a range of content including news, features, interviews, blogs, and industry reports.
WWW: LI profile
How do AI and other trends impact the way organizations manage and run their processes?
Organizations are being pushed to rethink processes from the ground up, not just automate steps, but redesign how work gets done, who does it, and how decisions are made. Whether its new technology like AI, changing customer expectations, or emerging regulatory requirements, modern businesses are under increasing pressure to be data-driven and resilient while remaining agile and human-centric – and that’s a fine balance!
Before, organizations designed processes upfront, documented them, and enforced compliance. Now, AI enables processes that learn and adapt in real time. Humans move from ‘doers’ to ‘orchestrators’ as AI changes roles, not just workflows. Employees supervise, validate, and fine-tune AI outputs Managers focus on outcomes, not micromanaging steps Process owners manage human–AI collaboration, not just SOPs.
What are the skills, techniques, behaviors, and attitudes that can help BPM practitioners create value for their organizations in 2026?
You simply can’t ignore AI and data literacy and understanding – it’s impacting pretty much all roles and industries. However, there is so much more to successful AI use and implementation than just technology. That’s where change management becomes essential.
In 2026, BPM practitioners create value less by drawing perfect process diagrams and more by shaping how work actually adapts, learns, and delivers outcomes. The role sits at the intersection of business, data, technology, and people. Great BPM practitioners are business translators and system designers who use data, AI, and human insight to continuously steer how work delivers value, rather than just documenting how it flows.
What are the best resources to learn those skills? (e.g. books, articles, courses)
PEX Network, of course! Joking aside, we pride ourselves on regularly publishing timely, high-quality content that not only keeps our audience up-to-date with the latest shifts in the industry but also supports learning and development. Of course, process excellence has long been associated with training and certifications, and this hasn’t changed even in the burgeoning AI/automation era. Methodologies like Lean Six Sigma and Agile still have value, but it’s about applying the core (and timeless) qualities of these approaches in a modern context.
Which skills are no longer relevant or not practically applicable yet (hype)?
Great question! This is where BPM maturity really shows in 2026. The biggest risk for BPM practitioners right now isn’t missing new skills, it’s over-investing in skills that no longer create value or that are still mostly hype.
Today, BPM practitioners lose value when they over-invest in heavyweight process documentation, rigid lifecycle models, centralized ‘process police’ governance, and tool-centric modeling skills, as these can’t keep up with fast-changing, data-driven work.
At the same time, much of the hype such as fully autonomous processes, AI-generated models as ground truth, digital twins of entire organizations, and perfect predictive BP remains impractical beyond narrow use cases due to data, trust, and regulatory limits.
The real risk is clinging to control, certainty, and ‘one best way’ thinking, rather than embracing adaptive, insight-driven, human-AI-orchestrated process management focused on outcomes and continuous learning.
Martin Holling
Industrial Engineer with 25+ years of experience in Business Process Management from operational implementation and improvement over QM, strategic development, process design and consultancy mainly in global corporations from small to more than 400.000 employees, focusing on Culture, people and continual improvement. Making use of broad experience in QHSE auditing, process documentation and project management implementation.
For further information about me and my ideas on BPM, you can have a look at both my LinkedIn profile and my website: https://living-processes.de/home-en/
WWW: LI profile
How do AI and other trends impact the way organizations manage and run their processes?
As organizations figure out that AI on its own does not result in a success, they will get more attention to their processes on how they are implemented and run in their business before they can successfully implement an AI initiative/solution. My hope is that there will be more focus on continual improvement and culture change in the processes to prepare for successful AI implementation.
What are the skills, techniques, behaviors, and attitudes that can help BPM practitioners create value for their organizations in 2026?
Soft skills, specifically in Change Management, People Integration and Moderating groups will separate the successful BPM practitioners from the ones that focus only on technical/technological skills and knowledge. These achieve much better process and implementation quality in the business that gives a fruitful basis for successful AI initiatives and even more efficient and effective processes that are much easier to automate.
What are the best resources to learn those skills? (e.g. books, articles, courses)
For sure there are courses and books out there on soft skills, but I think it is best to adapt your behavior by getting in touch with as many colleagues out there as possible. Go, get together with fellow BPM practitioners in active communities and learn from each other, might it be online or even better in personal meeting. Books can help to verify behavior and get initial ideas on what to change but meeting the people will get you to learn.
Which skills are no longer relevant or not practically applicable yet (hype)?
It is not that these skills are getting irrelevant, but process modelling and documentation will be more and more a thing that AI can do for you. You need to be able to understand it in depth and fine tune and correct the AI created process models and documents, but for example “translating” a process model from one notation to another one, might become an automated thing pretty soon.
Sandeep Johal
Sandeep is a Managing Director & Principal Consultant at Nano Business Technology with over 15 years of Business Process Management and Digital Transformation experience, specifically in enterprise wide system implementation process design, process improvement, strategic sourcing, capability uplift, strategy alignment, thought leadership in energy, utilities & resources; finance; and government bodies across Australia, New Zealand, Middle East, and North America
Sandeep’s consulting takes him to both national and international destinations including the Americas, Middle East, New Zealand and the UK. He is often invited to speak at national and international conferences and is regarded as a contributor to the Business Process Management body of knowledge. He holds a Masters in Information Technology (BPM), an honours in Business Management and a diploma in Mechanical Engineering.
WWW: Company website
WWW: LI profile
How do AI and other trends impact the way organizations manage and run their processes?
Organisations are feeling fatigued by the astonishing rate of AI progress and the pressure to keep up. Some have chosen to ‘watch this space’ before taking bold steps. Fortunately, solution vendors are aware of the AI race and are taking proactive steps to introduce the technology progressively.
2026 is often described as the year of the AI Agent. Trends point towards AI‑augmented process execution, where an AI Agent is constantly listening and contributing when required. Some predict that processes will eventually be AI Agent‑led and human‑augmented. Personally, given the rate of organisational adoption and the security implications, organisations are more likely to embrace a human‑led approach that is augmented by AI Agents.
Human‑led process execution will continue to involve automation. Humans will remain in charge of efficiency. AI Agents will continue to be integral to automation and efficiency, with the added capability of proactively addressing process improvements. Learning from these improvements will enable self‑correcting processes. Achieving this milestone will mark a true step towards intelligence in process management. Achieving this milestone will mark true intelligence in process management.
What are the skills, techniques, behaviors, and attitudes that can help BPM practitioners create value for their organizations in 2026?
AI literacy is no longer optional. Equipping BPM practitioners with AI basics and an understanding of the ever‑changing landscape of capabilities is essential. This allows practitioners to speak the language and understand what AI technology can and cannot do.
Foundational skills in process workshopping, problem definition, and modelling or visualisation are still relevant. In fact, the interest of major solution vendors such as Salesforce and SAP in acquiring process‑mining tools indicates that process visualisation and modelling remain highly relevant in 2026. Organisations still lack effective ways to bring together processes from disparate systems. BPM practitioners should view this as an opportunity to develop unifying mechanisms such as Process Architecture.
What are the best resources to learn those skills? (e.g. books, articles, courses)
One of the most interesting resources I’ve come across is the design of AI Agents in a visual studio that resembles traditional process‑modelling tools. AI Agent design platforms such as Zapier and Microsoft Copilot Studio employ drag‑and‑drop functionality to create agents, with options to connect to popular web services such as Gmail. There are heaps of video tutorials on YouTube about these platforms—well worth exploring.
For those interested in deep technical process design (for example, value stream mapping), a useful resource is the book Operational Excellence in Your Office: A Guide to Achieving Autonomous Value Stream Flow with Lean Techniques by Kevin J. Duggan and Tim Healey.
Which skills are no longer relevant or not practically applicable yet (hype)?
Unlike in previous years, I’m not aware of any BPM practitioner skill that is no longer relevant. Most skills remain essential, although some are applied differently. For example, creating the As‑Is of a process is often seen as wasteful. However, repositioning the As‑Is as a baseline validation for future improvement means that focused As‑Is detail is still required—just not exhaustive detail.
Emiel Kelly
I have been working “in BPM” for more than 25 years. Most of his time as a consultant and trainer at a BPM software and consulting organization. I helped all kind of companies in their BPM journey. From companies with 5 employees till companies with thousands of employees. From city councils, till investment companies and manufacturers of satellites.
Eight years ago I decided I want to make more impact on one company and joined an Insurance company (5 minutes cycling from my home). Of course I am still ‘doing BPM’ but with a much higher impact because I am part of the team now and fully responsible for the results of my implementations of ‘process things’. I can’t get away with leaving a slide deck behind, anymore 
As a hobby, I started my blog ‘Procesje.nl’ in 2011. The goal of this blog is to address the “nonsense” I run into in BPM world. Mainly brought with some irony, but always with the goal to help organizations make their processes perform better and stay away from the non value adding things.
WWW: https://procesje.blogspot.nl
WWW: LI profile
How do AI and other trends impact the way organizations manage and run their processes?
AI solves all problems! At least that’s what a lot of companies (at least the C people) think.
That’s nonsense of course. But with a BPM way of working in mind it can really help to improve things. On all levels of BPM.
On process design level AI can be a sparring partner to help you make clear
– the Why of a process?
– Useful KPI’s of the process
– What is needed to implement the process?
– What data is needed to check process performance?I’ve also seen AI that models processes. If that just leads to a picture of blocks and arrows, it has not much value. If it helps to create implementable workflows; yes!
On process execution level AI can execute some steps on it’s own or support the people in the process.
On case management level AI (if the data is available) can operate as some kind of flight control; keeping track of all the cases in the process and if they are still meeting their goals. If not, maybe AI can take some action or send out a warning.
On process improvement level AI can act as process mining on steroids; understand where bottlenecks arise, but more important what are the causes of those bottleneck, as bottlenecks are just symptoms of a bad process implementation.
What are the skills, techniques, behaviors, and attitudes that can help BPM practitioners create value for their organizations in 2026?
Still my number one is “strategic thinking”. BPM and processes are only a means. A means to solve the problems of customers. So always keep in mind if you are still solving the right problems. Help your company to implement useful processes. Help them make clear the why of the company an it’s processes.
It’s easy to dive in to process implementation very fast, but try to prevent that with your strategic view.
What are the best resources to learn those skills? (e.g. books, articles, courses)
Ask some generative AI. Pretty sure it will come up with my blog
Which skills are no longer relevant or not practically applicable yet (hype)?
My answer for many years has been high level modeling of processes. Of course those models are always right because they don’t tell the real story. Real processes are detailed. Very hard to catch in models. Happy that AI can help me now to really understand the dynamics of execution in a process. Having said that; without useful process data, AI lies to you. So I used it practically, but also had to apply a lot of common sense to not implement wrong improvements.
Guillermo Lopez
Working in the BPM Competence Center at the European Commission, I have spent the last eight years leading a team of experts to drive digital transformation and the modernization of EU institutions. I hold multiple certifications, including Professional Scrum Master, as well as specialized training in digital transformation, artificial intelligence, and process mining.
My core competencies include business process management, enterprise architecture, artificial intelligence, and agile methodologies. My mission is to help the EU deliver better services and outcomes to its citizens and stakeholders by leveraging state of the art BPM and EA technologies and methodologies.
I bring over 30 years of experience successfully leading BPM and EA projects across various domains and sectors—such as finance, retail, the public sector, and the environment—achieving significant improvements in efficiency, quality, and innovation.
WWW: LI profile
X: @GoreML
How do AI and other trends impact the way organizations manage and run their processes?
I don’t think AI will lead organizations; it will always be – or should always be – a technology under human supervision, with clear visibility into its reasoning and decision criteria. If multiple companies use the same kind of AI to shape their strategy, there’s a real risk they end up making very similar decisions and losing competitive differentiation. I’m also concerned about a “copycat” effect: erroneous strategies generated by a model being replicated uncritically across different organizations.
I don’t believe today’s generative AI will radically transform the world of processes and organizations, because it lacks deep context and doesn’t truly understand the world it operates in. When other kinds of AI emerge – like the family of approaches LeCun has proposed – that can build a solid representation of the environment and learn from it autonomously, then they may be able to lead and run truly autonomous enterprises. The current generation of models looks more like a powerful tool in the toolbox, not something that should play a leadership role.
Where I do see clear room for improvement is in case management: a constellation of agents helping you make better decisions and suggesting the most reasonable next steps to reach a given goal. AI can also add value in process mining analysis, process simulation, synthetic data generation, and similar tasks where its ability to explore scenarios and combine information is genuinely useful.
I’m particularly worried about three risks: removing the human-in-the-loop (HITL), starting from incorrect or biased input data, and the ultimate human responsibility for automated executions they may not fully understand. All of this makes me doubt that the current AI paradigm is the right path if it’s adopted as-is. On top of that, I see strong pressure to “move fast” and accelerate AI initiatives, and I think that’s a mistake: before making the whole organization “dance” to the tune of AI systems, we should first put in place strict governance, with clear rules on where, how, and under what constraints these models are used.
What are the skills, techniques, behaviors, and attitudes that can help BPM practitioners create value for their organizations in 2026?
In 2026, professionals will need to shift their mindset from a reactive stance to a clearly proactive one.
Aptitudes
• Ability to work with a far greater number of data sources (data fluency).
• Skill in interacting with and “questioning” AI systems with sound judgment (AI literacy).
• Significant strengthening of interpersonal and communication skills (soft skills).
• Deep understanding of the organization and its context: less of a “diagrammer” and more of an enterprise translator between business, technology, and people.
• Focus on responsible AI and governance: treating AI as a co pilot, not an oracle; demanding transparency, clear guardrails, and well defined accountability for AI influenced decisions and automations.
• Openness to change: viewing new tools (AI assistants, intelligent automation, unified BPM platforms) as leverage rather than threats and continuously updating one’s own methods.
• Customer and employee experience orientation: measuring success not only in cycle time or cost, but also in reduced friction for customers and frontline staff.
• Collaboration over control: moving away from a “central BPM police” model towards enabling process ownership in the business, with BPM acting as coach and backbone.Core skills and techniques
• Strong command of BPMN and DMN, which will remain essential, and the ability to review and refine AI generated models.
• Process mining and analytics: ability to formulate the right questions, interpret variants and bottlenecks, and propose concrete redesigns based on the findings.
• Automation and orchestration: knowledge of BPM engines, RPA, event driven architectures, and the ability to design flows with human in the loop as a central element, avoiding AI based black boxes.
• Simulation and experimentation: use of scenario simulation, what if analysis, and A/B testing to compare process designs and quantify impact before large scale implementation.
• Data and AI literacy: understanding what LLMs, ML models, and analytical models can and cannot do, how they depend on data quality, and where they fit in the BPM lifecycle (from documentation authoring through to decision support).
What are the best resources to learn those skills? (e.g. books, articles, courses)
• Academic–practitioner bridges for developing data fluency will remain very important, through initiatives such as bpm education or MultiProcessMining.
• It is also worth regularly following process mining and BPM trend blogs (for example, BOC Group, PrimeBPM, or PEX), as well as communities centered on commercial platforms (ARIS Community, Celonis, SAP Signavio, etc.).
• Another very good option is to follow leading voices in the field, such as Wil van der Aalst, Ian Gotts, or Jim Sinur.
• In addition, more and more university programs are emerging on BPM, the combination of BPM with AI, and process mining, such as some of the programs offered by the Universidad Internacional de La Rioja (UNIR), among others.
Which skills are no longer relevant or not practically applicable yet (hype)?
No longer relevant
• Static documentation as the main deliverable.
• Highly centralized BPM acting as a “process police” function.
• Modelling for the sake of modelling, with no clear link to real decisions or change.
• KPIs defined and maintained manually or without backing from operational data.
• Process discovery done only through workshops, without cross checking against execution data.
• Treating processes purely as technical problems, ignoring people and business context.
• Endless discovery and modelling sessions with no hypotheses and no measurable outcomes.Not really applicable yet (mostly hype)
• A fully autonomous enterprise with no transparency, no clear guardrails, and not well defined accountabilities.
• “In AI we trust” as a principle, delegating critical decisions to AI without questioning them.
• No one being accountable for what AI does: lack of an explicit framework for AI responsibility and accountability.
• Processes run without any visual representation that people can understand.
• Autopilots and black boxes “running the company”, without explainability or effective human oversight.
Matúš Mala
I’m Matúš, co-founder of the Process Academy, organizer of the BPM-Münich Meetup, podcast co-host of “The Process Philosophers” and an absolute BPM enthusiast.
WWW: LI profile
How do AI and other trends impact the way organizations manage and run their processes?
I think that by 2026, we should accept that AI is here to stay. The first disruption phase is over; AI is changing the way we work, prepare for meetings, communicate, think and live.
To me, it’s a new infrastructure technology similar to the internet in its early days, and now we have the opportunity to forget about FOMO (fear of missing out) and focus on real use cases.
I think that AI implementations by ‘end customers’ will slow down; they will no longer create new LLMs, RAGs, etc. without a target or purpose. Instead, they would focus on real improvements to their business processes, “How can AI help my core processes?”
On the other hand, I think there will be a ridiculous amount of new features in “tools/software”, creating “co-pilots” for almost everything, and I must say that I love it.
By the end of 2026, I think we will have much better tools and software solutions that will make it easier for us to create processes and solutions in our special BPM area.
What are the skills, techniques, behaviors, and attitudes that can help BPM practitioners create value for their organizations in 2026?
That’s a great question!
I truly think that we have a lot of great methodologies, techniques and more in BPM that will still be needed in 2026. The way we understand processes will not change. The opposite would be true: we are currently challenging new workflows in agentic orchestration, where requirements engineering and methodological questions arise, and the ‘old-school’ methods of understanding processes would be important here.
However, we should not cling to outdated methods; we should start to “refresh” our approach. Everything in our work is changing, so I think it’s extremely important that “old” experts and “new” newcomers develop updated BPM methodologies that will help in 2026 and beyond. We should work on questions such as: What kind of workshops do we need? What kind of structures do we need? Do we need new, ‘modern’ process landscapes? How can we improve requirement engineering? How should we describe processes (not only with BPMN)? And how can we spread BPM skills faster and more widely?
There are so many workflow tools, not only BPM tools, with diverse ways of creating and modelling workflows. I don’t think we will reduce them, so we have to adapt and make it easier to understand processes and create solutions using diverse tools, frameworks and more.
If I had to pick one skill: Flexibility would be key.
What are the best resources to learn those skills? (e.g. books, articles, courses)
That’s a difficult question.
The problem isn’t that we don’t have enough resources. Just go online or use an AI chatbot and you will find enough. The challenge here is that only people with intrinsic motivation do it. For those people, the form of knowledge is not that important; they want to learn and accept bad resources, difficult explanations, and so on. People learnt like this in the past and would continue to do so.
BPM, processes, data, AI… All of these topics are now important for everyone, for every employee. Now more than ever, it is important that everyone understands what AI is for, what they can and cannot do, and so on. The same applies to company governance, processes and data. These people are not usually intrinsically motivated to “learn” independently. It is therefore becoming increasingly important for companies to motivate them to learn, because the world is changing so quickly at the moment. Pure study is no longer enough; lifelong learning is essential.
I am not sure if we will see any improvements in the next couple of years, because normally companies don’t invest in these important topics, which is sad. They create some “learning paths” and short videos, but I just don’t see employees enjoying them. In fact, I think it’s worse than it was in the past. At least there were 2–3 days of workshops away from the office, and people were happy to learn and enjoy other places — it was a win-win situation. Currently, we just say, “Here are six 30-minute videos. Take a look…”
My advice: Until companies change their philosophy, find your favourite source and don’t push yourself: conferences, podcasts, your favourite YouTube channel, shorts, etc.
Which skills are no longer relevant or not practically applicable yet (hype)?
AI-only skills
For the last two years, we were somehow flying in the clouds, thinking that you don’t need anything but a prompt. You don’t even need to understand processes, data or anything else, just prompt.
Surely, people with less experience or technical knowledge can achieve more, but they need extremely high-level engineering skills to describe their “problem” or “solution”.
A poor process would be poor in AI as well.
Focusing only on prompting would not be that important anymore. A better understanding of problems and processes would be important. However, many technical disciplines would become less important. It is much easier now to create custom services and UIs. And it will improve even more. As with BPM, I think software engineers will become more “coordinating” agents. In the future, there will be fewer pure code solutions and more low-code or AI-engineered code solutions and models.
In short: A strong focus on one discipline (e.g. I am a Java programmer, I am prompt engineer, I am modeller …) is not future-oriented.
Morten Marquard
Morten Marquard has dedicated his entire professional journey to addressing the challenges faced by knowledge workers, including lawyers, social workers, and other professionals dealing with complex work processes. The struggle to navigate these processes efficiently while complying with ever-changing laws and regulations has been a persistent issue. Traditionally, compliance has relied on laborious reading and understanding of lengthy paper-based documents—a cumbersome task that often hinders productivity.
Recognizing the need for a transformative solution, Morten embarked on a mission to leverage technology for the benefit of knowledge workers, not only enhancing the efficiency and effectiveness of employees but also alleviating the burden of manual compliance checks and reducing stress levels.
Morten realized the limitations of using Business Process Model and Notation, BPMN, to streamline process digitalization as the rigidity of the processes failed to meet the requirements of end-users. It was during this critical juncture, approximately 15 years, that Morten collaborated with professor Thomas Hildebrandt, and together, they propelled the development of dynamic condition response graphs, DCR. This innovative approach has since been embraced by over 40 different customers, primarily in Denmark, with expanding reach into international markets such as Italy and the United Kingdom.
Morten’s journey exemplifies a commitment to pushing the boundaries of technology to empower knowledge workers, offering them a more streamlined and stress-free approach to managing their intricate work processes. The impact of his work extends far beyond national borders, contributing to a global shift in how organizations approach digitalization and compliance in the modern age.
WWW: LI profile
How do AI and other trends impact the way organizations manage and run their processes?
We are moving away from ‘Digital Theater’, where we just put PDFs on a screen, to the Agentic Stack. We use Generative AI to read the mess of regulations, but we don’t let it run the business. Why? Because LLMs are statistical; they guess! If you’re a student in Cambridge, a chatbot might say ‘yes’ to a beer, forgetting you’re at Cambridge Massachusetts, not England. In 2026, we manage processes by marrying LLMs for language with Symbolic AI for logic. This is the Business Operating System or Agentic AI: hardware-independent, sovereign, and 100% deterministic AI platform.
What are the skills, techniques, behaviors, and attitudes that can help BPM practitioners create value for their organizations in 2026?
Stop being a ‘Translator’ and start being a ‘Rule Architect.’ The most dangerous person in an organization is the one ‘building bridges’ between IT and Business. Bridges just keep the gap wide, and often widens it. We need to close the gap completely. The winning behavior in 2026 is Business Process Re-engineering (BPR) with a red marker. Michael Hammer: Don’t automate, obliterate. Don’t ‘digitize’ your old habits. If your process requires a person to manually type data into a CRM, don’t build an integration, kill the task! Practitioners must learn to empower business experts to own the logic directly through Declarative Process Modeling. We don’t need more ‘electronic’ paper; we need ‘Digital Twins’ of the organization where the logic is live, explainable, and hosted on European infrastructure.
What are the best resources to learn those skills? (e.g. books, articles, courses)
Stop reading generic “Success Stories” and start studying the intersection of Process Science and Computational Law. As Edsger W. Dijkstra famously warned, treating computers like humans is a sign of “professional immaturity.” We must stop pretending AI “thinks” or “understands” and start enforcing the formal logic our businesses depend on.
Real professions, like Law, Math, and Physics, developed specific languages precisely to avoid the ambiguity of “natural” language. Relying on the “vibe” of an AI is a step backward. For a practical deep-dive into how we fix this, look out for the upcoming book by Professor Thomas Hildebrandt and myself. It is the definitive guide to moving beyond “vibe coding” and into production-ready, rule-based engineering. We’ve been “too busy” in the trenches with our customers to finish it until now, but the era of “guessing” is over.
Which skills are no longer relevant or not practically applicable yet (hype)?
The post BPM Skills in 2026 (part 2) first appeared on BPM Tips.Retire the ‘Happy Path’ and stop ‘Building Bridges.’ The ‘Happy Path’ is a myth. As Professor Wil van der Aalst notes, 80% of cases follow their own unique paths. If you are still teaching ‘Lean’ flowcharts that break the moment reality hits, your skills are obsolete. People aren’t stupid; they deviate because they have to.
But the biggest ‘skill’ to unlearn? Bridge building. For years, we’ve hired ‘translators’ to sit between Business and IT. All they do is facilitate an expensive, digital game of telephone. The expert explains the law, the analyst writes a requirement, and the developer codes it. By the time it’s finished, the law has changed and the logic is lost in translation.
Also, stop the hype around RAG (Retrieval-Augmented Generation). Asking a chatbot to ‘read your manuals’ and guess an answer is irresponsible for Law, Finance, or GovTech. In 2026, the ‘Vibe Coding’ era is over for production. If your AI can’t provide a symbolic, explainable audit trail for its decisions, it’s just a toy. We don’t need ‘probably correct’ business processes; we need Compliance by Design.
In 2026, we don’t build bridges; we close the gap. The future belongs to the Business Operating System where the expert who knows the law is the one who defines the logic. IT should deliver the secure, sovereign infrastructure (Open Source and Kubernetes), but the business must own the execution. If you are still ‘translating’ requirements in 2026, you aren’t helping, you’re just slowing us down. Stand Tall Europe by letting the business take back the baton.