Applause, the global leader in digital quality and crowdsourced testing, has released its third annual State of Digital Quality in AI Survey. The findings reveal a major disconnect between rising investments in generative AI (Gen AI) and the adoption of essential quality assurance (QA) practices in software development.

With Gen AI applications and agentic AI rapidly evolving worldwide, ensuring rigorous testing throughout the software development lifecycle (SDLC) is crucial. These technologies, capable of autonomous decision-making without human intervention, introduce significant risks if not thoroughly vetted. Applause surveyed over 4,400 independent software developers, QA professionals, and consumers to explore AI use cases, tools, challenges, and user preferences.

AI Authority Trend: Rootstock AI Survey: 82 Percent of Manufacturers Boosting AI Budgets for 2025, Highlighting Need for AI-Ready ERP

AI Adoption Lags Behind Investment

“Given the massive investment in AI, we need to raise the bar on how we test and deploy new generative AI models and applications,” said Chris Sheehan, EVP of High Tech & AI at Applause. “Many organizations still struggle to integrate AI-powered productivity tools across the software development lifecycle. However, our global clients are ahead of the curve by embedding comprehensive AI testing measures earlier in development—from training models with high-quality datasets to implementing adversarial testing.”

Key Findings from the Survey

AI Testing Boosts Productivity, Yet Many Companies Are Slow to Integrate It

  • Over 50% of software professionals believe Gen AI tools significantly enhance productivity. About 25% reported a 25-49% boost, while 27% saw improvements between 50-74%.
  • However, 23% of developers said their integrated development environment (IDE) lacks embedded Gen AI tools like GitHub Copilot or OpenAI Codex, while 16% were unsure whether such tools were integrated.
  • Only 33% of respondents use red teaming, a critical practice that helps mitigate risks like inaccuracy, bias, and toxicity.
  • The top AI testing activities involving humans include prompt-response grading (61%), UX testing (57%), and accessibility testing (54%). Additionally, 41% of developers rely on domain experts to train industry-specific AI models.

Businesses Prioritize AI for Customer Experience but Struggle with Quality Issues

  • Over 70% of developers and QA professionals report that their organizations are actively developing AI applications.
  • Chatbots and AI-powered customer support tools lead AI adoption, with 55% of organizations working on these solutions. Meanwhile, 19% have started building AI agents.
  • In the past three months, 65% of users reported encountering issues with Gen AI responses, including lack of detail (40%), misinterpretation of prompts (38%), bias (35%), hallucinations (32%), incorrect information (23%), and offensive content (17%).
  • AI users are highly selective—30% have switched from one service to another, and 34% prefer different Gen AI tools for different tasks.

Additional Insights

  • Consumer demand for multimodal AI has surged. Nearly 78% of consumers now prioritize AI tools that can interpret multiple media types, up from 62% last year.
  • GitHub Copilot (37%) and OpenAI Codex (34%) remain the top AI-powered coding tools. However, their usage gap has narrowed compared to 2024.
  • QA professionals increasingly use AI for testing support. The top use cases include test case generation (66%), text generation for test data (59%), and test reporting (58%).

AI Authority Trend: Ciena Global Survey Explores Networking Needs for AI Era

Sheehan emphasized the role of human intelligence in AI development: “Enterprises leading the AI revolution recognize that human expertise remains critical. Every Gen AI use case requires tailored quality measures, from data training to real-world testing. As AI continues to shape our daily lives, delivering exceptional user experiences while mitigating risks must remain a top priority.”

About the AI Survey

The State of Digital Quality in AI Survey is part of Applause’s annual State of Digital Quality Report. Drawing insights from over 15 years of experience working with global enterprises and AI innovators, the report provides valuable guidance for organizations investing in AI and emerging technologies.

FAQs

1. Why is AI testing essential despite increasing investments?

AI investments are growing, but inadequate testing can lead to inaccurate, biased, or even harmful AI outputs. Quality assurance ensures reliability, safety, and user satisfaction.

2. What are the biggest AI challenges businesses face today?

Companies struggle with AI hallucinations, prompt misinterpretations, and biases. Many also lack integrated Gen AI tools in their development environments, slowing adoption.

3. How can businesses improve AI reliability and user experience?

Organizations should embed AI testing throughout development, utilize diverse training datasets, implement adversarial testing, and involve human evaluators to refine AI accuracy and fairness.

By prioritizing comprehensive AI testing, businesses can fully capitalize on their AI investments while ensuring trust, accuracy, and superior user experiences.

AI Authority Trend: 95 Percent of Contract Managers Are Open To Using AI, Per Recent CobbleStone Survey

To share your insights, please write to us at sudipto@intentamplify.com