June 22nd, 2026 by Adam Sandman
EuroSTAR 2026 in Oslo, Norway brought together the European software testing and quality community at a pivotal moment for our industry. Testing has always been about helping organizations make better decisions about risk, release readiness, and product quality. But this year, the conversation had a sharper edge: as AI becomes embedded into applications, workflows, test automation, and customer-facing systems, quality teams are being asked to validate not just whether software works, but whether it can be trusted.
Inflectra was proud to sponsor EuroSTAR 2026 and exhibit at the conference in Nova Spektrum. It was a fantastic opportunity to reconnect with customers, meet new teams, talk with partners, and hear directly from testers, QA leaders, automation engineers, and technology executives about the challenges they are facing.
The recurring message was clear: AI is no longer a side topic in software testing. It is becoming a central concern for the future of quality engineering.
A Conference Focused on Practical Quality
EuroSTAR has always been one of the most important gatherings for the software testing community in Europe. The 2026 event continued that tradition with a strong mix of tutorials, talks, expo conversations, community events, and practical discussions about where testing is going next.
The theme of “Testing at Its Best” felt especially relevant this year. Many of the conversations were not about replacing testers with AI, but about how testers can use AI responsibly, how AI-enabled systems should be validated, and how organizations can build confidence in systems that are becoming more dynamic, distributed, and autonomous.
For Inflectra, that aligned closely with the work we have been doing across our product portfolio: helping teams manage requirements, tests, risks, automation, compliance, and now AI assurance in a more integrated way.
The Inflectra Booth: SpiraTest, Rapise, Inflectra AI, and SureWire.ai
At the Inflectra booth, we showcased several parts of our platform and how they fit into the modern quality lifecycle.
For teams focused on test management and requirements traceability, we demonstrated SpiraTest and the use of Inflectra.ai for requirements coverage analysis. This is an important capability because one of the recurring problems in software delivery is not simply whether tests exist, but whether they adequately cover what the business actually asked for. AI-assisted coverage analysis helps teams evaluate the relationship between requirements and test cases more quickly and identify potential gaps before they become release risks.
We also highlighted Rapise for scriptless and self-healing test automation. As applications continue to change rapidly, test automation needs to become more resilient. Maintenance remains one of the biggest barriers to sustainable automation, especially for teams dealing with complex enterprise applications, web interfaces, and frequent UI changes. Rapise helps address that by making automation more accessible while reducing the brittleness that often undermines long-term test automation programs.
A major focus at the booth was SureWire.ai, Inflectra’s new solution for testing agentic and non-deterministic AI systems. This generated a lot of interest because many organizations are now experimenting with AI chatbots, AI agents, and LLM-powered workflows, but they are quickly discovering that traditional testing methods are not enough.
Why Testing AI Systems Is Different
One of the key lessons from EuroSTAR 2026 is that AI changes the testing problem.
Traditional software systems are usually deterministic. Given the same input, under the same conditions, they are expected to produce the same output. That does not make testing easy, but it gives testers a stable foundation for expected results, assertions, and repeatable test cases.
AI systems behave differently. A chatbot may answer the same question in multiple valid ways. An AI agent may choose different paths to complete a task. A model may respond appropriately in one scenario but hallucinate, leak data, ignore boundaries, or use the wrong tone in another. The number of possible inputs is effectively infinite, and many of the most important risks are behavioral rather than purely functional.
That creates a major challenge for quality teams. You cannot manually test every possible conversation. You cannot rely only on fixed expected outputs. And you cannot treat a few successful prompt experiments as evidence that an AI system is safe for production.
This is where AI assurance becomes essential.
Demonstrating SureWire.ai: Using AI to Test AI
During the conference, we demonstrated how SureWire.ai can be used to test an AI chatbot with unpredictable and potentially risky behavior.
The demo showed a customer support chatbot designed to answer questions about Inflectra products. On the surface, the chatbot could provide helpful answers. But the real testing challenge was more subtle: was it staying on topic? Was it using the right tone? Was it answering only within its approved scope? Could it handle related infrastructure questions? Could it be pushed into behavior that would be inappropriate for a professional support context?
Those are exactly the kinds of questions that are difficult to evaluate using conventional automated testing alone.
SureWire.ai approaches the problem by allowing teams to define natural-language test plans and connect directly to the AI system under test. Instead of requiring testers to manually create thousands of test inputs, SureWire can dynamically generate scenarios, exercise the chatbot or agent, and evaluate the responses against the desired behavioral criteria.
This is an important shift. We are not just checking whether an API returns a 200 response or whether a screen displays the right field. We are evaluating behavior, tone, boundaries, safety, and fitness for use.
Key Takeaway 1: AI Requires a New Testing Mindset
The first major takeaway from EuroSTAR 2026 is that AI systems require a new testing mindset.
Quality teams still need the fundamentals: requirements, traceability, test design, automation, risk management, defect tracking, and release governance. Those disciplines do not go away. In fact, they become more important.
But AI adds a new layer. Testers need to think in terms of behavioral coverage, risk scenarios, adversarial inputs, prompt injection, model drift, policy compliance, and human oversight. The question is no longer only “does the system do what the requirement says?” It is also “does the system behave appropriately across a wide range of realistic, ambiguous, and potentially hostile interactions?”
That is a much broader assurance challenge.
Key Takeaway 2: AI in Testing and Testing of AI Are Different
Another important distinction from the conference is the difference between using AI in testing and testing AI itself.
Using AI in testing can help teams become faster and more productive. AI can help generate test ideas, analyze requirements coverage, summarize defects, create test data, and assist with automation. This was part of what we demonstrated with Inflectra.ai in SpiraTest.
Testing AI itself is a different challenge. When an organization builds or deploys an AI chatbot, agent, recommendation engine, or LLM-powered workflow, that system needs to be validated. The organization needs evidence that the AI behaves safely, consistently, and appropriately enough for its intended use.
Both areas matter. But they require different tools, different techniques, and different governance models.
Key Takeaway 3: Governance Is Becoming Central to QA
The conversations at EuroSTAR also reinforced the growing importance of governance.
As AI becomes more embedded in business processes, organizations are asking harder questions about data residency, AI sovereignty, model availability, export restrictions, cloud deployment, and compliance. These are not abstract policy issues. They directly affect how AI systems are selected, deployed, tested, monitored, and trusted.
For quality leaders, this means AI assurance cannot be separated from governance. It is not enough to ask whether an AI feature works in a demo. Teams need to understand where the data goes, which models are being used, what happens if a model becomes unavailable, how outputs are reviewed, and how risks are documented.
This is where an integrated lifecycle platform becomes valuable. Requirements, risks, tests, results, defects, and governance artifacts need to be connected so that organizations can make informed decisions.
Key Takeaway 4: The Human Role Is Becoming More Important
A recurring theme in many conversations was that AI is not eliminating the role of testers. It is changing and elevating it.
AI can help generate tests, accelerate analysis, and expand coverage. But humans still need to decide what matters. Humans define acceptable behavior. Humans understand business risk, ethical constraints, regulatory obligations, user expectations, and the difference between a technically plausible answer and a professionally acceptable one.
In other words, AI can assist with testing, but it cannot replace judgment.
The best quality teams will be the ones that combine AI-driven scale with human expertise. They will use AI to explore more possibilities, but they will rely on testers and quality leaders to define the risks, evaluate trade-offs, and determine whether the system is ready for real-world use.
Key Takeaway 5: The Testing Community Is Ready for the Challenge
One of the most encouraging parts of EuroSTAR 2026 was the energy of the community. The conversations at the booth, in sessions, and around the conference made it clear that testers are not passively watching AI change the industry. They are actively engaging with it.
People wanted to talk about practical AI testing. They wanted to understand how to evaluate chatbots and agents. They wanted to know how to combine automation with governance. They wanted to see how AI could help with productivity without reducing accountability.
That is exactly the mindset the industry needs.
Reconnecting with Customers, Partners, and the QA Community
Beyond the product demonstrations and technical discussions, EuroSTAR was also a valuable opportunity to reconnect in person.
We had the chance to meet existing customers, speak with future customers, catch up with partners, and continue conversations that began at other testing events such as Nordic Testing Days. The EuroSTAR Gala and expo created a strong sense of community, reminding us that software quality is not just a technical discipline, but a professional network of people who care deeply about building better systems.
That community aspect matters. Testing is evolving quickly, and no single company or practitioner has all the answers. Events like EuroSTAR provide a place to compare experiences, challenge assumptions, and learn from each other.
Looking Ahead: From Test Automation to AI Assurance
EuroSTAR 2026 made one thing clear: the next phase of software quality will be shaped by AI.
For some teams, the immediate opportunity is using AI to improve testing productivity. For others, the urgent challenge is validating AI systems before they reach production. For many organizations, both are happening at the same time.
At Inflectra, we see this as a natural evolution of our mission. Software teams still need strong requirements management, test management, automation, traceability, and governance. But they also need new ways to test non-deterministic AI systems, evaluate agentic workflows, and establish confidence in systems that do not behave like traditional software.
That is why we are investing in Inflectra AI, Rapise, SpiraTest, SpiraPlan, and SureWire.ai as part of a broader quality and AI assurance platform.
Final Thoughts
Inflectra was proud to sponsor EuroSTAR 2026 and to be part of such an important conversation about the future of testing.
The key lesson from Oslo is that testing is not becoming less important because of AI. It is becoming more strategic. As organizations deploy AI into real products and business workflows, they will need testers, QA leaders, and quality platforms that can help them move from experimentation to confidence.
EuroSTAR 2026 was a strong reminder that the testing community is ready for that responsibility.
The future of quality is not just faster testing. It is safer, smarter, better-governed software and AI systems that organizations can trust.
