The potential transformative impact of Artificial Intelligence (AI) on science, education, work and society is critical to the future of employee assistance. Our survey study was designed to better understand where, how and why EAPs are employing AI at their organizations and vendor businesses. We wanted to understand what is working, what is not working and why. Additionally, we asked about ethical and operational concerns. To this end, we conducted an online self-report survey in early 2025. With the support of the 8 major professional groups in the industry, the researchers issued an international call for participation. We collected data from 222 survey respondents including a mix of embedded internal staff, external vendor and hybrid model programs situated in a diverse range of workplaces and practice settings. Voices of EAPs were represented in the study from Argentina, Australia, Belgium, Bulgaria, Canada, England, Germany, Greece, Hong Kong, India, Japan, Luxemburg, Malaysia, New Zealand, Nigeria, Poland, United Kingdom, United States, Slovenia, South Africa, South Korea, and Taiwan. Results are presented the subgroup of 14 respondents who were members of the Employee Assistance Professionals Association of Australasia and a few others who also lived in the same countries of Australia and New Zealand (n=3). This group of 17 was compared with the remaining 203 participants. Key findings: Current use of AI technology is low today with just 38% currently using AI tools (with 85% of this group starting to do so within just the last two years). However, a large majority (78%) of the total respondents believe that AI will play an increasing role in the future of EAP service delivery. Overall, 32% of EAPs are actively exploring how to add AI tools, 30% are interested in adding AI resources if certain challenges are resolved and 32% don't know or are against AI altogether and prefer human solutions instead. Of 14 specific applications of AI in EAP, most had similar levels of interest to add to services in the future. Other results reflect a profile of both positive and negative attitudes toward AI technology. The average rating on a 0 to 10 scale was: 5.8 for feeling positive about AI and 4.6 for feeling negative about AI. However, these were inversely rated (r = -.51) with individuals such that the typical respondent was not ambivalent and had a mostly positive or mostly negative emotional tone. Three general groups were proposed to describe the study sample: The Connected (mostly positive; 27%); the Critical (mostly negative; 18%) and the Cautious (moderate for both emotions; 45%).