Functional testing just crossed a meaningful threshold. For years, the path from manual to automated testing has demanded significant time, technical expertise, and ongoing maintenance. Automation has remained constrained to a subset of tests, while most of the coverage stays manual, slow, and increasingly unsustainable.
Today, autonomous testing lets teams run existing manual tests without writing a single automation script, freeing up time and expanding coverage in ways that simply weren’t practical before.
With OpenText Core Software Delivery Platform Functional Testing 26.2, we’re introducing capabilities that change how teams think about test automation and AI quality: the Autonomous Tester, AI response evaluation, and a new way to work with model-based testing (MBT) units. Rather than a collection of incremental updates, each of these capabilities reflects a deliberate shift toward autonomous, AI-driven quality, where testing adapts to change and becomes accessible to a broader range of teams.
The Autonomous Tester: Run your own manual tests, automatically
Until now, moving from manual to automated testing meant investing significant time and expertise to convert test cases into scripts, build frameworks, and maintain them as your application changes. That investment has always been the biggest barrier to scaling test coverage.
The Autonomous Tester removes that barrier almost entirely. Assign your existing manual tests to the Autonomous Tester runner and set them to run automatically in a test suite, no scripting or conversion required. Tests run unattended, and you get detailed run evidence and results including agent actions, step status, and recordings.
Your team stays in control throughout. While the Autonomous Tester handles execution, QA engineers and testers remain the decision-makers. This is automation that works alongside your team, not a black box that replaces their judgment. Every run produces a full audit trail of agent actions, step-level status, and recordings, so your people always have the visibility they need to act with confidence.
Here’s what teams gain from day one:
- Automate existing manual tests instantly without rewriting scripts or rebuilding frameworks
- Accelerate release cycles by eliminating manual test execution overhead
- Scale test coverage without the overhead of manual execution
- Keep humans in the loop with full run transparency through agent actions, step-level status, and recordings for every run
- Resolve issues sooner using OpenText Aviator to analyze test and suite run results for faster issue resolution
Note: This feature requires an Aviator license.
AI response evaluation: Test your own AI
As more organizations deploy AI agents, coding assistants, and LLM-powered features in their products, a practical question has emerged—how do you test something that doesn’t produce deterministic outputs? Standard functional testing approaches don’t map cleanly to GenAI behavior, where responses vary, context matters, and traditional pass/fail logic falls short.
AI response evaluation in codeless tests addresses this directly. Using the new Evaluate GenAI response step, you can validate the behavior and reliability of GenAI and LLM-based intelligent systems before they reach production. You’re able to configure evaluation methods, metrics, and pass levels to enable confidence-based evaluation for non-deterministic responses.
The feature is built for subject matter experts, not just automation engineers:
- Validate GenAI behavior without any AI, ML, or model-tuning expertise — so the people who know the business logic best can own the evaluation
- Replace binary yes/no and pass/fail logic with confidence-based validation designed for the unpredictable nature of non-deterministic responses
- Understand results immediately with readable and explainable outputs Maintain enterprise-grade governance with built-in auditability across every evaluation run
- Adapt evaluation criteria as your AI applications grow in complexity, keeping quality standards current without rebuilding from scratch
Note: This feature requires an Aviator license.
More precision in model-based testing
Version 26.2 also brings a meaningful improvement to how teams work with model-based testing. You can now create unit steps directly in the new Steps tab, just as you would in a manual test. No need to link to a separate manual test.
What does this mean for your MBT workflows?
- Describe detailed unit actions directly in the Steps tab without separate manual test dependencies
- Surface step-level detail in the Manual Runner and Manual Run report
- Pinpoint MBT test failures faster with granular visibility into exactly where things broke down
Taken together, these updates reduce the friction that has historically slowed MBT adoption. Teams can build more detailed tests without jumping between tools, and when something fails, they spend less time working out where things broke down. The result is a tighter feedback loop and more confidence in every test run.
Digital lab
Whether you’re upgrading your mobile test stack to take advantage of Appium’s latest improvements or simply trying to spend less time hunting for the right device, these updates get out of your way and let you focus on what matters.
- Run tests on Appium v3 with full support for the latest major release of the widely adopted mobile automation framework. The lab now includes an embedded Appium v3 server, so there’s nothing extra to install or maintain.
- Switch versions without rebuilding by using the appiumVersion capability to choose v3 in your test configuration. Existing test assets carry over as-is.
- See device availability at a glance with total item counts now displayed alongside each filter option in the lab portal. Know what’s available before you select it, with no extra clicks required.
OpenText Functional Testing Lab for Mobile and Web continues to evolve alongside the tools and frameworks your teams rely on, so your testing infrastructure doesn’t become the bottleneck.
Testing that keeps pace
Ready to see what version 26.2 can do for your team?
Together, these capabilities reflect where enterprise testing is heading. More autonomous, more intelligent, and more accessible to the people who understand the business logic best, not only the engineers. Whether you’re just getting started with automation or looking to extend your quality strategy into AI-powered applications, version 26.2 gives you real, production-ready tools to move faster and test smarter.
Explore the full release details in the OpenText Core Software Delivery Platform Functional Testing 26.2 What’s New Help Center.
Want to test it out first? Start a free trial today. 
Miss the OpenText Functional Testing 26.1 release in our enterprise DevOps platform? Visit the What’s Your Version? web page.
Explore the OpenText DevOps Cloud Community to connect with other users and share what you’re building!
