Artificial Intelligence (AI) is rapidly reshaping team collaboration in workplaces. Meanwhile, women only represent 22% of the global AI workforce, raising questions about whose perspectives drive AI design. This dual reality makes the stakes high: without critical attention, AI may entrench existing gendered dynamics; but with deliberate design, it may open new avenues for equity. Through in-depth interviews with 30 AI professionals (22 women), our work both confirms gendered challenges in male-dominated teams and offers a novel contribution: how those who directly experience these dynamics envision AI’s role in mitigating them. These practitioner-informed design visions reveal AI’s potential of offering multi-level support and empowering women to navigate these teams, and its risks of reinforcing stereotypes and surveillance. We call on the HCI community to explore this emerging design space for equitable human-AI teaming while critically attending to gendered power dynamics.

Trust plays a critical role in both effective teamwork and successful integration of AI into human-AI teams (HATs), which are becoming increasingly interconnected to tackle large-scale problems. As these HATs grow in complexity, so do the challenges of building, maintaining, calibrating, and aligning trust across HATs, as trust is dynamic, socially influenced, and capable of spreading and evolving over time. To understand the underexplored question of how trust and distrust can spread across HATs, we conducted a laboratory experiment with two interconnected HATs (N=40) distributed across the U.S., each comprising two humans and one AI collaborating over a remotely piloted aircraft system (RPAS) conducting simulated reconnaissance tasks. Our findings show that distrust toward the AI teammate spread more easily than trust, both within and across HATs. While intra-HAT (dis)trust in the AI developed through direct interaction, inter-HAT (dis)trust was more vulnerable to word-of-mouth distrust. Additionally, participants’ trust in the (dis)trust spreader, as well as their team-level trust, reflected in-group favoritism and out-group bias. This work presents the first empirical investigation into trust and distrust contagion across HATs, expanding the current understanding of trust dynamics in multi-agent and multi-team systems. It identifies potential mechanisms driving trust and distrust transfer between interconnected HATs and explores the individual, team, and multi-team level impacts. In doing so, we lay the groundwork for developing theories on trust contagion in multi-agent, multi-team systems, providing valuable insights for future research and design strategies to optimize human-AI collaboration in complex environments.

Artificial Intelligence (AI)-mediated collaboration has been a core research interest in CSCW and HCI in recent decades. AI is being embedded in teams as a teammate, known as human-AI teamwork. A critical aspect to this and any type of teamwork is monitoring: the ability to track and be aware of both team- and individual-level actions. As AI continuously evolves from a passive tool into a supportive collaborator capable of playing a role in teams, more research is needed to investigate human perceptions of AI teammates' monitoring. More work is needed on when and how AI can function as a supportive collaborator rather than a surveillant (in regard to monitoring) to enhance team coordination and performance. Drawing on 22 interviews, we examined how humans perceive AI monitoring when AI behaves as a collaborative teammate for team support, focusing on perceptions of AI monitoring at both team and individual levels. Four tensions have been identified, namely: performance versus privacy, support versus control, fairness versus empathy, and team-level and individual-level trade-offs. We also propose feasible design principles for human-AI teaming designers to optimize AI monitoring for effective team support, maximizing its unique advantages while mitigating potential risks, further cultivating more effective and trustworthy human-AI collaboration.

With rapid advancements in large language models (LLMs), generative AI (GenAI) is transforming people's life and work across various domains. Unlike previous AI technologies that are often feminized, most of these GenAI tools are non-gendered, potentially preventing users from applying gender stereotypes. However, GenAI's use of natural language can evoke social perceptions including gender attribution, making it susceptible to gender associations. Using two online experiments, we explored how GenAI's removal of gender could mitigate individuals' gender stereotypes toward it, and how certain linguistic cues could trigger gender stereotypes even if it is non-gendered. We found that the removal of AI's gender did mitigate gender stereotypes toward it, but only to an extent. Additionally, gendered linguistic cues such as politeness, apologies, and tentative language (or lack thereof) could trigger gender stereotypes toward non-gendered AI. We contribute to HCI and CSCW research by providing a timely investigation into individuals' gender stereotypes toward GenAI agents, and one of the first to examine the profound impact of language style on users' perceptions and attributions of social characteristics to GenAI. Our findings shed light on the purposeful and responsible design of GenAI that prioritizes promoting gender equality, thereby ensuring that technological advancements align with evolving social values.

AI increasingly assumes complex roles in Human-AI Teaming (HAT). However, communication and trust issues between humans and AI often hinder effective collaboration within HATs, highlighting a need for effective human-centered team training, an area significantly understudied. To address this gap, we interviewed eSports athletes and team-based, competitive gamers (N=22), a group experienced in HATs and team training, about their HAT team training needs and desires. Through the lens of Quantitative Ethnography (QE), we analyzed their insights to understand preferred team training strategies and the desired roles of AI within these strategies, considering the varying levels of human expertise. Our findings reveal a strong preference across all expertise levels for cross-training, which is training in other teammate roles, to improve perspective taking and coordination in HATs. Less experienced participants prefer structured procedural training, while experts favor self-correction methods for growth. Additionally, participants desired that AI act as a companion, with beginners and intermediates valuing AI's functional roles, and experts seeking AI in a coaching role. Among the first to emphasize human-centered team training in HATs, this study contributes to CSCW/HCI research by revealing varied preferences for training and AI roles, emphasizing the need to tailor these aspects to team dynamics and individual skills for better outcomes in HATs.

The success of human-AI teams (HATs) requires humans to work with AI teammates in trustful ways over a certain time period. However, how trust evolves and changes dynamically in response to human-AI team interactions is generally understudied. This work explores the evolvement of trust in HATs over time by analyzing 45 participants' experiences of trust or distrust in an AI teammate prior to, during, and after collaborating with AI in a three-member HAT. Our findings highlight that humans' expectations of AI's ability, integrity, benevolence, and adaptability influence their initial trust in AI before collaboration. However, this initial trust can be maintained or revised through the development of situational trust during collaboration in response to the AI teammate's communication behaviors. Further, the trust developed through collaboration can impact individuals' subsequent expectations of AI's ability and their collaborations with AI. Our findings also reveal the similarities and differences in the temporal dimensions of trust for AI and human teammates. We contribute to CSCW community by offering one of the first empirical investigations into the dynamic and temporal dimension of trust evolvement in HATs. Our work yields insights into the pathways to expanding the methodological toolkit for investigating the development of trust in HATs, formulating theories of trust for the HAT context. These insights further inform the effective design of AI teammates and provide guidance on the timing, content, and methods for calibrating trust in future human-AI collaboration contexts.

People often apply gender stereotypes toward computerized agents. Rather than challenging these stereotypes, modern AI technologies controversially use them in creating AI agents that underline stereotypical gendered roles. This approach thus further reinforces the often male-dominated societal gender norms and disenfranchises women and gender non-binary individuals. While this issue has raised concerns in the HCI and CSCW communities, still little is known regarding how to mitigate the negative impacts of embedding gender stereotypes in AI agents. In this paper, we propose an eXplainable AI (XAI) approach to mitigating individuals' gender stereotypes toward AI agents. We conducted an online video vignette experiment with 350 participants randomly assigned to one of the eighteen conditions of a 3 (gender of the agent: woman, man, gender-neutral) x 3 (task gender: feminine, masculine, neutral) x 2 (presence or absence of AI explanation) between-subjects design. Our findings indeed suggest that XAI helped participants avoid applying gender stereotypes toward gendered AI agents, by increasing their understanding of how the agent came to its decision and decreasing their rating of the agent's humanlikeness. We contribute to CSCW research by providing a timely investigation into individuals' gender stereotypes toward the state-of-the-art AI agents and by advancing the empirical understanding of the cognitive processes and mechanisms underlying these gender stereotypes. We also demonstrate how eXplainable AI can effectively suppress the application of social characteristics (i.e., gender stereotypes) toward AI agents by disrupting the said cognitive processes. Insights from this study can inform how future AI technologies should be designed to create a progressive gender reality that will gradually reshape humans' experience and ingrained gender ideologies.

In the coming years, Artificial Intelligence (AI) will be applied as a teammate that works alongside and collaborates with humans. Prior research in teaming and CSCW has shown that teammates have the ability to change the thoughts and behaviors of each other through simple interactions in a process known as social influence. However, to date, research has yet to identify the social influence that AI teammates could have in these human-AI teams, which has led to a limited understanding of how AI teammates will change the behaviors of their human teammates. To remedy this gap, we conduct a mixed-methods study (N=33) with young individuals to explore how humans could behaviorally adapt and perceive their behavioral adaptation due to interaction with an AI teammate. Qualitative results report that perceived three unique stages they had to experience for the social influence of their AI teammate to lead to adaptation (i.e., perceiving a sense of control, identifying a technological or performative justification, and gaining first-hand experience). Quantitative results validate and illustrate the results of this perceived process, as results show that participants adapted their behaviors to complement the behaviors of different types of AI teammates. This study contributes to the CSCW/HCI field by developing an initial understanding of AI teammates' social influence in human-AI teams, which will be a pivotal design and research consideration in future efforts.

Newly-formed teams often encounter the challenge of members coming together to collaborate on a project without prior knowledge of each other's working and communication styles. This lack of familiarity can lead to conflicts and misunderstandings, hindering effective teamwork. Derived from research in social recommender systems, team recommender systems have shown the ability to address this challenge by providing personality- derived recommendations that help individuals interact with teammates with differing personalities. However, such an approach raises privacy concerns as to whether teammates would be willing to disclose such personal information with their team. Using a vignette survey conducted via a research platform that hosts a team recommender system, this study found that context and individual differences significantly impact disclosure preferences related to team recommender systems. Specifically, when working in interdependent teams where success required collective performance, participants were more likely to disclose personality information related to Emotionality and Extraversion unconditionally. Drawing on these findings, this study created and evaluated a machine learning model to predict disclosure preferences based on group context and individual differences, which can help tailor privacy considerations in team recommender systems prior to interaction.

As AI becomes more advanced in natural language processing, the research on AI's communication in teaming environments is getting more attention in CSCW/HCI. Even though AI's communication serves as an essential part of facilitating team coordination and shaping team outcomes, the impact of AI's communication modality on human-AI teamwork is still understudied. Using a mixed-design experiment and follow-up group interviews with 100 participants, we investigate the impact of AI's communication modality, one of AI's most essential communication characteristics, on team coordination and team outcomes in two different human-AI team compositions, human-human-AI teams, and human-AI-AI teams. Our findings highlight the trade-offs between AI's verbal communication and visual communication, which inspire two design recommendations on how to apply AI's verbal and visual communication to support human-AI coordination effectively. Our study generates an initial understanding of the role that AI's communication will have in ensuring team effectiveness in human-AI teams in the CSCW/HCI field.

Explanation of artificial intelligence (AI) decision-making has become an important research area in human–computer interaction (HCI) and computer-supported teamwork research. While plenty of research has investigated AI explanations with an intent to improve AI transparency and human trust in AI, how AI explanations function in teaming environments remains unclear. Given that a major benefit of AI giving explanations is to increase human trust understanding how AI explanations impact human trust is crucial to effective human-AI teamwork. An online experiment was conducted with 156 participants to explore this question by examining how a teammate’s explanations impact the perceived trust of the teammate and the effectiveness of the team and how these impacts vary based on whether the teammate is a human or an AI. This study shows that explanations facilitate trust in AI teammates when explaining why AI disobeyed humans’ orders but hindered trust when explaining why an AI lied to humans. In addition, participants’ personal characteristics (e.g., their gender and the individual’s ethical framework) impacted their perceptions of AI teammates both directly and indirectly in different scenarios. Our study contributes to interactive intelligent systems and HCI by shedding light on how an AI teammate’s actions and corresponding explanations are perceived by humans while identifying factors that impact trust and perceived effectiveness. This work provides an initial understanding of AI explanations in human-AI teams, which can be used for future research to build upon in exploring AI explanation implementation in collaborative environments.

Accurate detection of gait event intention and sending it to lower limb exoskeleton (LLE) is the key to achieve active rehabilitation. Most existing surface electromyography (sEMG)-based gait event intention detection methods suffer from insufficient generalization and complex detection. In this paper, we propose a novel approach for detecting gait event intention using a single sEMG sensor. The gait event intention is obtained by detecting the peak activity of the rectus femoris during the stance period. First, the root mean square (RMS) features are extracted from the sEMG data of the rectus femoris. Then, the data groups composed of the RMS features are smoothed and all extreme points are calculated. Finally, the midstance (MSt) events are discovered when the latest maximum point satisfies the preset condition. The experimental results of three different gait speeds showed that the proposed approach could adapt to different walking speeds and maintain a high detection accuracy of gait event intention detection. This study provides a convenient and reliable detection approach for gait research of LLE.

Non-native speakers (NNS) often face challenges gaining the speaking floor in conversations with native speakers (NS) of a common language. To help NNS to contribute more, we developed a conversational agent that opens up the speaking floor either automatically, after NS have taken a certain number of consecutive speaking turns, or manually, upon NNS request. We compared these automatic and manual agents to a control condition in a laboratory study in which one NNS collaborated with two NS using English as a common language. Participants (N=48) communicated over video conferencing from separate locations in a research institution to collaborate on three survival tasks. Based on data gathered from the experiments, the automatic agent encouraged NNS to participate more, which previous studies had attempted but failed to achieve. Excerpts from group discussions further showed the crucial role of the automatic agent on NNS participation. Interview results suggested that while NNS appreciated the automatic agent's help to participation, NS perceived the agent's interruption as unfair because they thought all members were speaking equally, which was not the case. The mismatch in their perceptions further emphasizes the need to intervene, and we provide design implications based on the results.

A growing organizational trend is to utilize ad-hoc team formation which allows for teams to intentionally form based on the member skills required to accomplish a specific task. Due to the unfamiliar nature of these teams, teammates are often limited by their understanding of one another (e.g., teammate preferences, tendencies, attitudes) which limits the team's functioning and efficiency. This study conceptualizes and investigates the use of a teammate information-sharing recommender system which selectively shares interpersonal recommendations between unfamiliar teammates (e.g., "Your voice may be overshadowed by this teammate when making decisions...") to promote teammate understanding. Through a mixed-methods approach involving 105 participants working on actual unfamiliar teams, this study explores how presentation elements such as anonymity and explanations influence system perceptions and how anonymity influences team outcomes. Results indicate that anonymizing recommendations was associated with worse team measures, particularly team satisfaction and team cohesion. Qualitative results shed light on why team members perceived privacy concerns and team benefits associated with using the system. We contribute to CSCW through a better understanding of how to support unfamiliar teams, the conceptualization and empirical investigation of a novel teammate information-sharing recommender system, and foundational design recommendations associated with such a system.

Recently, AI is integrating into teams to collaborate with humans as a teammate with the goal of achieving unprecedented team outcomes. Much of the coordination between humans and AI teammates relies on human-AI communication, which is challenging due to AI's limitations on natural language communication. Thus, it is essential to identify and develop effective communication strategies for AI teammates in human-AI teams to facilitate the coordination process. Through interviews with 60 participants who collaborated with an AI teammate in a multiplayer online game, in this paper, we explore communication strategies that humans expect AI teammates to apply to support human-AI coordination and collaboration in dyadic teaming environments, and how the AI teammate's communication can impact teaming processes. Our findings highlight four communication strategies AI teammates should apply to support their coordination with humans in dyadic teaming environments. We also find that AI teammates' proactive communication with humans could facilitate the development of human trust and situation awareness, whereas AI lacking such proactive communication is often not perceived as a teammate. Our study extends the current CSCW/HCI research on human-AI communication in teaming environments by shedding light on how communication should be structured in dyadic human-AI teams for effective and smooth collaboration.

Multiparty collaboration using a common language is often challenging for nonnative speakers (NNS). Conversation can move forward rapidly, with terms and references unfamiliar to NNS often going unexplained because NNS do not request clarification due to cognitive overload or face concerns. Language difficulties may further lead to NNS having a low level of participation in a conversation, which could be a loss for multilingual teams. To help NNS resolve potential confusions due to unfamiliar language use without risking face concerns, we created a conversation agent that asked clarification questions intended to help NNS follow and participate in multiparty conversations. We conducted a within-subjects laboratory experiment with 17 triads of 2 NS and 1 NNS, who performed a series of collaborative tasks under three conditions: a) no agent, b) a high-level agent that resembles a NNS with good command of English, and c) a low-level agent that resembles a NNS with poor English skills. Results suggest that NS made significantly more clarifications in both agent conditions than without an agent. In the high-level agent condition, NNS reported an increase in understanding after the agent's interruption and spoke significantly more. Further, NNS evaluated their communication competence in English highest in the low-level agent condition and lowest in the control condition. Our findings suggest several directions to improve the tool to better facilitate multilingual multiparty communication.

Collaborating using a common language can be challenging for non-native speakers (NNS). These challenges can be reduced when native speakers (NS) adjust their speech behavior for NNS, for example by speaking more slowly. In this study, we examined whether the use of real-time speech rate feedback (a speech speedometer) would help NS monitor their speaking speed and adjust for NNS accordingly. We conducted a laboratory experiment with 20 triads of 2 NS and 1 NNS. NS in half of the groups were given the speech speedometer. We found that NS with the speech speedometer were significantly more motivated to slow down their speech but they did not actually speak more slowly, although they made other speech adjustments. Furthermore, NNS perceived the speech of NS with the speedometer less clear, and they felt less accommodated. The results highlight the need for tools that create scaffolding to help NS make speech accommodations. We conclude with some design ideas for these scaffolding tools.