General Discussion

Which aproach do you prefer? * Anthropic to watermark all AI text (presumably to stop slop)* Google lets you turn off the visible watermark on AI-made images, videos, and music in Gemini and Flow. The funny thing is the end result is the same in both cases.Just because we either see or do not see the watermark is it still going to be there ( Invisible SynthID and C2PA provenance still stay behind for verification) So..can you see me? 

I'm a non-corporate user with a Tosca trial-license issue and I can't find a way to reach support.Background: I finished AS1 and AS2 and started the TDS1 course a few months ago, but when I wanted to continue today Tosca Commander could no longer load a valid license from the cloud license server.Attempt 1: I checked under https://support-hub.tricentis.com/open?id=training_license_request and it shows two training license requests with state ‘Deployed’ with the latest updated on "2026-05-14” and the [Request Cloud Training License]  button is greyed out with the tooltip “User already requested too many training license requests”. (There is no delete button or any other option either.)Attempt 2: So I sent an email. Got an email back saying "we have discontinued the option to open Academy cases (former: incidents) via e-mail. Please log into the Tricentis Support Hub and connect with the Virtual Agent to find a solution” Fine.Attempt 3: I go to the support hub, find the virtual agent and

I asked AI to generate a regression suite.It understood the assignment.Maybe a little too well.It generated:✅ 427 login tests✅ 189 password reset tests✅ 96 variations of "user clicks the button"And...One payment test.Apparently, losing millions is less important than making sure the Login button works on a Tuesday.😂This is the part of AI-generated testing that doesn't get enough attention.AI is fantastic at generating tests.It doesn't automatically understand risk.It doesn't know: which feature makes your company money which workflow customers use the most which bug will wake up your CTO at 2 AM which release would make LinkedIn explode with angry posts That's still our job.The future of Quality Engineering isn't spending all day writing test scripts.It's reviewing, prioritizing, and teaching AI what actually matters.I'll happily let AI generate 1,000 tests.But before I merge them, I'm asking one question:"Which of these would I be sad to delete?"If the answer is "almost none"..

AI tools & productivity in a nutshell:Save lists → never openBookmark resources → never useSubscribe to newsletters → never read(Your “Saved for later” folder is basically a digital graveyard at this point.) Who actually improves? One strong source + a ruthless filter.Be honest… how big is your digital graveyard right now? 

Hi Team,Recently completed courses is transferred to new mail id. But for completed courses AS1 and AS2 , completion certificates are not available which was earlier available. Also at course level i can see course being completed, But module level entire course is showing as not completed.I tried to connect to support , but getting permission related issue.Kindly help in this regard. 

Everyone is talking about AI writing test automation.But here's the question I think we're asking too infrequently:Does AI actually understand your product? Check the below video with common "AI tells" that appear in generated tests—patterns that reveal the model is writing syntactically correct code without understanding the business it's supposed to validate. We look at four common pathologies:Coverage blindspotsThe mocking trapBrittle assertionsMechanics over meaning AI is an incredible productivity tool, but passing tests don't necessarily mean you're testing the right things.If you're using AI to generate Playwright, Cypress, Selenium, or unit tests, these are the signs to watch for. I'd love to hear your experience: Have you noticed any AI tells in the tests it generates?#AI #TestAutomation #SoftwareTesting #QualityEngineering #Playwright #Selenium #QA #SDET #SoftwareEngineering #GenerativeAI https://www.youtube.com/watch?v=qeIFv5uBM0M  

Hello ,I have completed the course and taken all quizzes and final exam. But certificate is not sent to my email. could you please check?thanks,Shruthi

Hi @PolinaKr, I'm having the same issue. I completed the "Gen AI for Testers" course by Rahul, including all quizzes and the final exam, but I don't see any certificate/badge on my profile (checked "Badges" section — it's empty). My profile: shiftsync.tricentis.com/members/maham-masood-21241. Could you please check and grant it? Thanks, Maham

HI, What’s UserID i need to give before starting the Quiz.? 

Hi TeamI have completed the “AI Fundamentals (Generative AI for Testers)” course and successfully completed all 5 quizzes. However, even after submitting the results, I have not yet received the completion certificate either in my profile or in my registered email.Could you please check and help resolve this issue? 

I saw this meme and… yeah, this is exactly what working with AI feels like right now 😅You finally get comfortable with one update…maybe even feel a little confident using it…And then:“New update”“Newest update”“Tomorrow’s update”At some point you just sit there like:“Wait… am I even using the right version anymore?” 😂I’ve caught myself going down the rabbit hole of trying to keep up with everything…and honestly, it’s exhausting.Lately, I’ve been approaching it differently:Instead of chasing every new release,I focus on understanding what actually changes and how it impacts my work.Because the truth is:Most updates don’t change everything… but they do change something that matters.And that’s where the real value is.Curious how others are handling this pace—Are you trying to keep up with every update, or just focusing on what’s relevant?#AI #TechLife #ContinuousLearning #SoftwareTesting #QualityEngineering

Hello!sorry to bother you all… it looks like I’m not enabled to use official support channels. Unfortunately, I missed the Detecting and Preventing LLM Hallucinations webinar, which could be very interesting for my work. Is there any way to get the recording?Thank you so much!

With a normal API or UI you test deterministic systems — same input, same output, every time. AI agents break that contract completely. The same prompt can produce different reasoning paths, different tool call sequences, and different outputs across runs. That non-determinism is the core challenge. Ηοw would you test this system? A question that sparks the skills needed for testers in the new era! 

Ever been ghosted by a JOB? Check the resource below https://www.youtube.com/shorts/xDIf0YFC1UI

A malicious file labeled “2026_Salary_Increase.pdf.exe” was opened by a large portion of employees — including the Incident Response team. Let that sink in. This isn’t about lack of intelligence. It’s about how easily human psychology can be exploited:Curiosity (salary increase)Urgency (what if I’m missing out?) Trust in familiar file naming patterns Key takeaways for teams: • File extensions still matter — .pdf.exe is a red flag• Security awareness training must be continuous, not one-off• Even security teams need realistic phishing simulations• Technology alone won’t save you — behavior will The real vulnerability isn’t the system. It’s the assumption that “this wouldn’t fool me.” When was the last time your organization tested this?#CyberSecurity #Phishing #SecurityAwareness #InfoSec #QA #RiskManagement 

A few years ago, I was proud of a UI automation suite I built.Hundreds of tests.Clean reports.Everything green.Until production started failing.Users reported errors we had never seen.But our UI tests were still passing.So I went deeper.Instead of clicking through the UI, I started testing the APIs directly.That’s when things became obvious:• The API returned inconsistent data• Some responses were missing critical fields• Error handling was completely broken• Edge cases were never validatedThe UI was masking all of it.It handled errors silently.It retried requests.It hid failures behind “friendly messages.”Our UI tests were verifying the illusion of stability.Not the system itself.That changed how I approach testing.Now I start with APIs:→ Validate contracts→ Break requests intentionally→ Test edge cases early→ Then verify UI behaviorBecause if the foundation is broken…Testing the surface won’t help.Do you start testing from the UI or the API?

It’s easy to jump too far in the future and imagine what AI will look like and how it may help us do all the things.However, in reality, we have to take steps towards a future that (may) look like that.What tedious, boring or repetitive tasks are you actually using AI for? (or plan to in the near future)

Hi everyone, I’m James, a pentester focused on web application and API security. I spend most of my time exploring vulnerabilities and real attack paths. For the past few years, I’ve been building ZeroThreat AI, an automated pentesting tool with exploit validation and agentic AI testing capabilities. I joined the community to share ideas, learn from others, and discuss how security testing is evolving from manual to automated, and then with agentic AI tools. Ciao!!!!

I once tested a password reset feature. Everything worked.Emails sent correctly.Tokens expired properly. But out of curiosity I tried something strange.I requested 100 password resets in 10 seconds. The system happily sent them all.That meant anyone could spam users with reset emails. Not a typical functional bug.But definitely a security and abuse issue.Sometimes testing means asking unusual questions like:"What happens if someone abuses this feature?" QA is not just about correctness.It's about resilience. Have you ever tested rate limits intentionally? 

One of the biggest lessons in my QA career came from a failed automation suite. We had hundreds of tests.Great coverage.Fast execution.But production still had bugs. Why?...........................Because our tests verified features, not user behavior. Real users behaved differently.They navigated unpredictably.They refreshed pages.They switched devices. Automation eventually improved when we started designing tests around user journeys instead of components. Automation should mirror how real people use the product.Do your tests reflect real user behavior?

Early in my automation career I believed something dangerous:"If automation passes, the system is safe."One release proved me wrong.All tests were green.CI pipeline passed.Deployment happened.Minutes later support reported users couldn't complete checkout.The issue?Our automation mocked the payment service.The real integration had changed.Automation verified the mock, not the system.That moment changed how I design tests.Now I always ask:Are we testing reality?Or just a simulation?Automation should reduce risk.But only if it tests real system behavior.Have you ever seen automation give a false sense of security? 

Hi All,My name is Deepak Behera. I’m a Quality Engineering professional with around 19 years of experience in software testing, with a strong focus on test automation and AI‑driven QA practices. My background spans telecom and complex enterprise systems, where I’ve worked as a hands‑on QA engineer and quality leader, building robust automation suites and improving test strategies end‑to‑end. I hold a BE in Computer Science and an MBA in Information Systems, and I’m currently scaling my expertise towards QE Architect and managerial roles while also founding Vim‑Ana (www.vim-ana.org) , a startup in the UAV/drone space where disciplined testing and reliability are critical.In this community, I’m looking to both share and learn: sharing practical insights on test automation frameworks, API testing, AI‑assisted testing, and QA leadership, and learning from others on how they are evolving their quality strategies, tooling, and career paths.My LinkedIn: https://www.linkedin.com/in/hideepak/

Compare the Tricentis AI testing with other AI products in the market.Provide the advantages over other tools in the market for the automation testing using Tricentis.