Testing AI with AI: Navigating the Challenges of QA

Robert Salesas
Leapwork

AI sure grew fast in popularity, but are AI apps any good?

Well, there are some snags. We ran some research recently that showed 85% of companies have integrated AI apps into their tech stack in the last year. Pretty impressive number, but we also learned that many of those companies are running head-first into some issues: 68% have already experienced some significant problems related to the performance, accuracy, and reliability of those AI apps.

If companies are going to keep integrating AI applications into their tech stack at the rate they are, then they need to be aware of AI's limitations. More importantly, they need to evolve their testing regiment.

The Wild Wild West of AI Applications

That AI apps are buggy isn't necessarily a damnation of AI as a concept. It simply draws attention to the reality that AI apps are being managed within complex, interconnected systems. Many of these AI apps are integrated into sprawling tech stack ecosystems, and most AI tools in their current form don't exactly work perfectly out of the box. AI applications require continuous evaluation, validation, and fine-tuning to deliver on expectations.

Without that validation process, you risk stifling the effectiveness of AI apps with bugs and security vulnerabilities (security risks were one of the most commonly flagged issues for AI applications). Ultimately, that means the company doing the integration just becomes exposed to system failures, decreased customer satisfaction, and reputational damage. And considering how reliant the world will likely soon be on AI, that's something every business should aim to avoid.

Fixing AI … with AI?

Ironically, the answer many companies seem to have settled on for fixing their testing inefficiencies is AI-augmented testing. We found that 79% of companies have already adopted AI-augmented testing tools, and 64% of C-Suites trust their results (technical teams trust even more at 72%).

Is that not a bit paradoxical? Why fix AI with more AI?

In the right context, AI-augmented testing tools can be that second set of eyes (long live the four-eyes principle) to vet the shortcomings of AI systems with rigorous, unbiased reviews of performance. The reason you would use AI-augmented testing is to gauge how well generative AI deals with specific tasks or responds to user-defined prompts. They can compare AI-generated answers versus predefined, human-crafted expectations. That matters when AI models so often hallucinate nonsensical information.

You can imagine the many linguistic permutations for asking an AI chatbot, "Do you offer international shipping?" A response needs to be factually right regardless of how the question was asked, and that's where AI-augmented testing tools shine in automating the validation process for variables.

Do We Need Human QA Testers?

There's just one outstanding question: What happens to the human QA testers if everyone starts using AI-augmented testing?

The short answer to this question? They'll still be around, don't you worry, because over two-thirds (68%) of C-Suite executives we've spoken to have said they believe human validation will remain essential for ensuring quality across complex systems.  Actually, 53% of C-Suite executives told us they saw an increase in new positions requiring AI expertise. Fancy that ...

There's a good reason why humans won't disappear from QA teams. AI isn't perfect, and that extends to testing. Some testing tools can do things like self-healing scripts where the AI adjusts a test in line with minor app changes, but they can't handle the complexity of most real-world applications without any human supervision. We have AI agents, but they don't have agency. Autonomous testing agents can't just suddenly decide independently to test your delivery app to check whether your pizza orders are going through.

All of which is to say that some degree of human validation will be needed for the foreseeable future to ensure accuracy and relevance. Humans need to be there to decide what to automate, what not to automate, and how to create good testing procedures. The future of QA isn't about replacing humans but evolving their roles. Human testers will increasingly focus on overseeing and fine-tuning AI tools, interpreting complex data, and bringing critical thinking to the testing process.

AI offers huge amounts of promise, but this promise created by adoption must be paired with a vigilant approach to quality assurance. By combining the efficiency of AI tools with human creativity and critical thinking, businesses can ensure higher-quality outcomes and maintain trust in their increasingly complex systems.

Robert Salesas is CTO of Leapwork

Related Links

Hot Topics

The Latest

The Real Enterprise AI Risk Isn't Hallucinations

Ask most IT leaders about their biggest concern with AI and you'll hear the same answer: hallucinations ... Today, however, the conversation has shifted ... As organizations move beyond chatbots and experiments, they are increasingly deploying AI agents that perform multi-step tasks. These systems retrieve documents, query databases, call APIs, generate reports, write code, and make recommendations. The issue is not whether the model can reason. The issue is whether the organization can see, verify, and govern the decisions being made along the way ...

Beyond Buy vs. Build: The New Playbook for Data Pipelines

While organizations want to take control of their telemetry, building telemetry pipelines from scratch can be a very daunting, complicated task, even when leveraging open-source standards like OpenTelemetry. It requires specialized knowledge across distributed systems, data engineering, and security. This fragmented approach across systems causes higher operational costs; it puts a strain on resources and reduces efficiency as teams have to work with different interfaces and processes ...

The Corporate Network Is Fading - Here's What Replaces It

For decades, enterprise networks were designed around a simple assumption: work happened inside the office. Applications lived in centralized data centers, employees connected through internal infrastructure, and security focused on protecting the perimeter that surrounded everything ... But the way organizations operate today bears little resemblance to that environment. Cloud platforms host critical applications, employees connect from homes and airports as often as they do from offices, and partners collaborate through shared systems that exist far beyond corporate walls. In short, the corporate network no longer resembles the environment it was designed to protect ...

IT Leaders Must Establish a Network Data Architecture Practice

As an analyst who researches how IT organizations design, build, and operate their networks, I find that network data is a constant source of pain. Network teams struggle with data quality, fragmentation, authority, access, and trust. And these issues undermine everything they try to do. Here are the numbers: Only 45% of network teams are completely confident in the accuracy of their network source of truth, which documents the intent of their network ...

Data Center Outages Are Decreasing but More Costly

The 2026 Global Data Center Survey from Uptime Institute reveals an industry navigating workforce constraints, escalating outage expenses, even as rising costs remain the top concern for management teams ...

Observability Has a Physical Layer Now: What AI Infrastructure Teaches IT Operations

The next observability gap may not be in the code. It may be under the rack. That sounds strange until you think about how AI incidents actually feel in the middle of an investigation ... The application dashboard may be accurate. It may also be stopping at the wrong boundary. AI systems depend on software, but they also depend on a dense physical stack: racks, power paths, thermal margin, maintenance activity and, in many environments, liquid cooling. Those physical dependencies can change slowly before they look like a software incident ...

TLS Certificate Expiration Is Becoming an Observability Problem

Certificate expiration is the rare outage you can see coming. Every TLS certificate carries the date it stops working, so the moment it will begin breaking connections is knowable in advance. That's what makes an expired certificate such a frustrating way to lose a service. What's changing now is how often that date comes around ...

Enterprise Cloud Observability: Choosing the Right Starting Point

Enterprises operate different combinations of workloads across cloud, hybrid and multicloud environments. For business-critical workloads, teams need to consider monitoring and observability early so they can detect health issues, investigate failures, and understand operational impact. Organizations place workloads on cloud platforms based on a combination of technical requirements, economics, existing dependencies, organizational standards, and business priorities. Their monitoring priorities therefore depend on what they operate and where those systems run. Those priorities will not look the same for every organization ...

AI Is Making Incident Management More Complex - Bad Data Makes It Harder

Top-performing businesses prioritize data-driven decision making, enabling leaders to move from intuition and gut feel towards evidence-based judgment. But that judgment is only sound when the data underpinning decisions is accurate. With incident management, data accuracy is particularly important. Long-term revenue, customer trust, and operational stability depend on high-quality data that enables teams to quickly identify and address the root cause of major incidents. Against this backdrop, governance becomes a critical endeavor to ensure the right data drives the right action ...

MEAN TIME TO INSIGHT Podcast - Episode 26: Network Compliance

In MEAN TIME TO INSIGHT Episode 26, Shamus McGillicuddy, VP of Research, Network Infrastructure and Operations, at EMA discusses network compliance ... 

Testing AI with AI: Navigating the Challenges of QA

Robert Salesas
Leapwork

AI sure grew fast in popularity, but are AI apps any good?

Well, there are some snags. We ran some research recently that showed 85% of companies have integrated AI apps into their tech stack in the last year. Pretty impressive number, but we also learned that many of those companies are running head-first into some issues: 68% have already experienced some significant problems related to the performance, accuracy, and reliability of those AI apps.

If companies are going to keep integrating AI applications into their tech stack at the rate they are, then they need to be aware of AI's limitations. More importantly, they need to evolve their testing regiment.

The Wild Wild West of AI Applications

That AI apps are buggy isn't necessarily a damnation of AI as a concept. It simply draws attention to the reality that AI apps are being managed within complex, interconnected systems. Many of these AI apps are integrated into sprawling tech stack ecosystems, and most AI tools in their current form don't exactly work perfectly out of the box. AI applications require continuous evaluation, validation, and fine-tuning to deliver on expectations.

Without that validation process, you risk stifling the effectiveness of AI apps with bugs and security vulnerabilities (security risks were one of the most commonly flagged issues for AI applications). Ultimately, that means the company doing the integration just becomes exposed to system failures, decreased customer satisfaction, and reputational damage. And considering how reliant the world will likely soon be on AI, that's something every business should aim to avoid.

Fixing AI … with AI?

Ironically, the answer many companies seem to have settled on for fixing their testing inefficiencies is AI-augmented testing. We found that 79% of companies have already adopted AI-augmented testing tools, and 64% of C-Suites trust their results (technical teams trust even more at 72%).

Is that not a bit paradoxical? Why fix AI with more AI?

In the right context, AI-augmented testing tools can be that second set of eyes (long live the four-eyes principle) to vet the shortcomings of AI systems with rigorous, unbiased reviews of performance. The reason you would use AI-augmented testing is to gauge how well generative AI deals with specific tasks or responds to user-defined prompts. They can compare AI-generated answers versus predefined, human-crafted expectations. That matters when AI models so often hallucinate nonsensical information.

You can imagine the many linguistic permutations for asking an AI chatbot, "Do you offer international shipping?" A response needs to be factually right regardless of how the question was asked, and that's where AI-augmented testing tools shine in automating the validation process for variables.

Do We Need Human QA Testers?

There's just one outstanding question: What happens to the human QA testers if everyone starts using AI-augmented testing?

The short answer to this question? They'll still be around, don't you worry, because over two-thirds (68%) of C-Suite executives we've spoken to have said they believe human validation will remain essential for ensuring quality across complex systems.  Actually, 53% of C-Suite executives told us they saw an increase in new positions requiring AI expertise. Fancy that ...

There's a good reason why humans won't disappear from QA teams. AI isn't perfect, and that extends to testing. Some testing tools can do things like self-healing scripts where the AI adjusts a test in line with minor app changes, but they can't handle the complexity of most real-world applications without any human supervision. We have AI agents, but they don't have agency. Autonomous testing agents can't just suddenly decide independently to test your delivery app to check whether your pizza orders are going through.

All of which is to say that some degree of human validation will be needed for the foreseeable future to ensure accuracy and relevance. Humans need to be there to decide what to automate, what not to automate, and how to create good testing procedures. The future of QA isn't about replacing humans but evolving their roles. Human testers will increasingly focus on overseeing and fine-tuning AI tools, interpreting complex data, and bringing critical thinking to the testing process.

AI offers huge amounts of promise, but this promise created by adoption must be paired with a vigilant approach to quality assurance. By combining the efficiency of AI tools with human creativity and critical thinking, businesses can ensure higher-quality outcomes and maintain trust in their increasingly complex systems.

Robert Salesas is CTO of Leapwork

Related Links

Hot Topics

The Latest

The Real Enterprise AI Risk Isn't Hallucinations

Ask most IT leaders about their biggest concern with AI and you'll hear the same answer: hallucinations ... Today, however, the conversation has shifted ... As organizations move beyond chatbots and experiments, they are increasingly deploying AI agents that perform multi-step tasks. These systems retrieve documents, query databases, call APIs, generate reports, write code, and make recommendations. The issue is not whether the model can reason. The issue is whether the organization can see, verify, and govern the decisions being made along the way ...

Beyond Buy vs. Build: The New Playbook for Data Pipelines

While organizations want to take control of their telemetry, building telemetry pipelines from scratch can be a very daunting, complicated task, even when leveraging open-source standards like OpenTelemetry. It requires specialized knowledge across distributed systems, data engineering, and security. This fragmented approach across systems causes higher operational costs; it puts a strain on resources and reduces efficiency as teams have to work with different interfaces and processes ...

The Corporate Network Is Fading - Here's What Replaces It

For decades, enterprise networks were designed around a simple assumption: work happened inside the office. Applications lived in centralized data centers, employees connected through internal infrastructure, and security focused on protecting the perimeter that surrounded everything ... But the way organizations operate today bears little resemblance to that environment. Cloud platforms host critical applications, employees connect from homes and airports as often as they do from offices, and partners collaborate through shared systems that exist far beyond corporate walls. In short, the corporate network no longer resembles the environment it was designed to protect ...

IT Leaders Must Establish a Network Data Architecture Practice

As an analyst who researches how IT organizations design, build, and operate their networks, I find that network data is a constant source of pain. Network teams struggle with data quality, fragmentation, authority, access, and trust. And these issues undermine everything they try to do. Here are the numbers: Only 45% of network teams are completely confident in the accuracy of their network source of truth, which documents the intent of their network ...

Data Center Outages Are Decreasing but More Costly

The 2026 Global Data Center Survey from Uptime Institute reveals an industry navigating workforce constraints, escalating outage expenses, even as rising costs remain the top concern for management teams ...

Observability Has a Physical Layer Now: What AI Infrastructure Teaches IT Operations

The next observability gap may not be in the code. It may be under the rack. That sounds strange until you think about how AI incidents actually feel in the middle of an investigation ... The application dashboard may be accurate. It may also be stopping at the wrong boundary. AI systems depend on software, but they also depend on a dense physical stack: racks, power paths, thermal margin, maintenance activity and, in many environments, liquid cooling. Those physical dependencies can change slowly before they look like a software incident ...

TLS Certificate Expiration Is Becoming an Observability Problem

Certificate expiration is the rare outage you can see coming. Every TLS certificate carries the date it stops working, so the moment it will begin breaking connections is knowable in advance. That's what makes an expired certificate such a frustrating way to lose a service. What's changing now is how often that date comes around ...

Enterprise Cloud Observability: Choosing the Right Starting Point

Enterprises operate different combinations of workloads across cloud, hybrid and multicloud environments. For business-critical workloads, teams need to consider monitoring and observability early so they can detect health issues, investigate failures, and understand operational impact. Organizations place workloads on cloud platforms based on a combination of technical requirements, economics, existing dependencies, organizational standards, and business priorities. Their monitoring priorities therefore depend on what they operate and where those systems run. Those priorities will not look the same for every organization ...

AI Is Making Incident Management More Complex - Bad Data Makes It Harder

Top-performing businesses prioritize data-driven decision making, enabling leaders to move from intuition and gut feel towards evidence-based judgment. But that judgment is only sound when the data underpinning decisions is accurate. With incident management, data accuracy is particularly important. Long-term revenue, customer trust, and operational stability depend on high-quality data that enables teams to quickly identify and address the root cause of major incidents. Against this backdrop, governance becomes a critical endeavor to ensure the right data drives the right action ...

MEAN TIME TO INSIGHT Podcast - Episode 26: Network Compliance

In MEAN TIME TO INSIGHT Episode 26, Shamus McGillicuddy, VP of Research, Network Infrastructure and Operations, at EMA discusses network compliance ...