ESEIW 2025
Sun 28 September - Fri 3 October 2025

This program is tentative and subject to change.

The rapid adoption of Large Language Models (LLMs) in the engineering of software applications, such as customer service chatbots, has brought significant benefits, but has also posed substantial risks. Generation of biased, inappropriate, or harmful responses are among the potential problems that can arise when using LLM as a COTS, connecting its chat to the user interface of a software application. This paper brings emerging results of an exploratory study that compares commercial guardrail frameworks to evaluate their ability to retain unappropriate content during a chat conversation. We empirically evaluate three guardrail frameworks — LLM Guard, Llama Guard, and OpenAI Moderation — against two datasets of toxic and offensive content. Results show that improvements are still needed, since the assessed guardrails frameworks achieved high accuracy for one of the datasets (more than 90%) but underperformed in other metrics, showing that toxic or dangerous content could still be delivered for users if this is deployed in a chatbot, for instance. We hope these results assist researchers and practitioners in selecting appropriate guardrails to improve harm mitigation in LLM-based applications.

This program is tentative and subject to change.

Fri 3 Oct

Displayed time zone: Hawaii change

15:40 - 17:00
Responsible and Inclusive AI in Software EngineeringESEM - Emerging Results and Vision Track / ESEM - Journal First Track / ESEM - Registered Reports Track / ESEM - Technical Track / at Kaiulani I
Chair(s): Italo Santos University of Hawai‘i at Mānoa
15:40
16m
Talk
Invisible Risks, Visible Code: A Vision for Understanding Ethical Debt in AI-Based Coding
ESEM - Emerging Results and Vision Track
Dina Salah City St George’s, University of London
15:56
16m
Talk
"Is It Responsible?" Emerging Results on Comparing Guardrails for Harm Mitigation in LLM-enhanced Software Applications
ESEM - Emerging Results and Vision Track
Manoel Veríssimo dos Santos Neto Universidade Federal de Goiás (UFG), Mohamad Kassab Boston University, USA, Arlindo Galvão Universidade Federal de Goiás, Valdemar Vicente Graciano Neto Universidade Federal de Goiás (UFG), Edson OliveiraJr State University of Maringá
16:12
16m
Talk
Trust, Transparency, and Adoption in Generative AI for Software Engineering: Insights from Twitter Discourse
ESEM - Journal First Track
Manaal Ramadan Basha The University of British Columbia, Gema Rodriguez-Perez The University of British Columbia
16:28
16m
Talk
Toward Inclusive AI-Driven Development: Exploring Gender Differences in Code Generation Tool Interactions
ESEM - Registered Reports Track
Manaal Ramadan Basha The University of British Columbia, Ivan Beschastnikh The University of British Columbia, Gema Rodriguez-Perez The University of British Columbia, Cleidson de Souza Universidade Federal do Pará
16:44
16m
Talk
Beyond Binary Moderation: Identifying Fine-Grained Sexist and Misogynistic Behavior on GitHub with Large Language Models
ESEM - Technical Track
Tanni Dev Wayne State University, Sayma Sultana Wayne State University, Amiangshu Bosu Wayne State University
Hide past events