Testing for accessibility is a significant step when developing software, as it ensures that all users, including those with disabilities, can effectively engage with web and mobile applications. While automated tools exist to detect accessibility issues in software, none are as comprehensive and effective as the process of user testing, where testers with various disabilities evaluate the application for accessibility and usability issues. However, user testing is not popular with software developers as it requires conducting lengthy interviews with users and later parsing through large recordings to derive the issues to fix. In this paper, we explore how large language models (LLMs) like GPT 4.0, which have shown promising results in context comprehension and semantic text generation, can mitigate this issue and streamline the user testing process. Our solution, called Reca11, takes in auto-generated transcripts from user testing video recordings and extracts the accessibility and usability issues mentioned by the tester. Our systematic prompt engineering determines the optimal configuration of input, instruction, context and demonstrations for best results. We evaluate Reca11's effectiveness on 36 user testing sessions across three applications. Based on the findings, we investigate the strengths and weaknesses of using LLMs in this space.

Artificial Intelligence (AI) techniques, especially Large Language Models (LLMs), have started gaining popularity among researchers and software developers for generating source code. However, LLMs have been shown to generate code with quality issues and also incurred copyright/licensing infringements. Therefore, detecting whether a piece of source code is written by humans or AI has become necessary. This study first presents an empirical analysis to investigate the effectiveness of the existing AI detection tools in detecting AI-generated code. The results show that they all perform poorly and lack sufficient generalizability to be practically deployed. Then, to improve the performance of AI-generated code detection, we propose a range of approaches, including fine-tuning the LLMs and machine learning-based classification with static code metrics or code embedding generated from Abstract Syntax Tree (AST). Our best model outperforms state-of-the-art AI-generated code detector (GPTSniffer) and achieves an F1 score of 82.55. We also conduct an ablation study on our best-performing model to investigate the impact of different source code features on its performance.

The adoption of Large Language Models (LLMs) is reshaping software development as developers integrate these LLMs into their applications. In such applications, prompts serve as the primary means of interacting with LLMs. Despite the widespread use of LLM-integrated applications, there is limited understanding of how developers manage and evolve prompts. This study presents the first empirical analysis of prompt evolution in LLM-integrated software development. We analyzed 1,262 prompt changes across 243 GitHub repositories to investigate the patterns and frequencies of prompt changes, their relationship with code changes, documentation practices, and their impact on system behavior. Our findings show that developers primarily evolve prompts through additions and modifications, with most changes occurring during feature development. We identified key challenges in prompt engineering: only $21.9 \%$ of prompt changes are documented in commit messages, changes can introduce logical inconsistencies, and misalignment often occurs between prompt changes and LLM responses. These insights emphasize the need for specialized testing frameworks, automated validation tools, and improved documentation practices to enhance the reliability of LLM-integrated applications.

Despite the availability of numerous automatic accessibility testing solutions, web accessibility issues persist on many websites. Moreover, there is a lack of systematic evaluations of the efficacy of current accessibility testing tools. To address this gap, we present the first mutation analysis framework, called Ma11y, designed to assess web accessibility testing tools. Ma11y includes 25 mutation operators that intentionally violate various accessibility principles and an automated oracle to determine whether a mutant is detected by a testing tool. Evaluation on real-world websites demonstrates the practical applicability of the mutation operators and the framework’s capacity to assess tool performance. Our results demonstrate that the current tools cannot identify nearly 50% of the accessibility bugs injected by our framework, thus underscoring the need for the development of more effective accessibility testing tools. Finally, the framework’s accuracy and performance attest to its potential for seamless and automated application in practical settings.