New research from the Notre Dame-IBM Tech Ethics Lab, in partnership with All Tech Is Human, shows that the increasing use of artificial intelligence (AI) in the workplace is not only creating new types of jobs but also changing how employees experience their work. The study was based on workshops with mid-level professionals and senior leaders, and it found that different people in different roles and with different identities often experience AI adoption in very different ways. Both groups noted that speed and productivity have become more important in the workplace, but they also said that overall standards have dropped. As a result, "good enough" work is becoming more common. Additionally, AI has introduced a range of new tasks that are critical for integrating AI into the workplace, but these tasks are rarely included in official job descriptions, performance reviews, or company metrics. These tasks are called "glue work," and they include checking and reviewing AI-generated outputs, translating high-level strategies into practical actions, coordinating across different departments, and training employees. Middle managers, in particular, have taken on a lot of this extra work. Some participants in the study mentioned that this new kind of labor is often shouldered more by women throughout the organization. Meanwhile, senior leaders questioned whether middle management would still be needed in the future in AI-driven companies. They also worried about "distributed deskilling," which refers to the gradual loss of opportunities for employees to develop skills and knowledge through informal means like mentoring, knowledge sharing, and training programs. The workshops revealed a growing conflict between the perspectives of middle managers and senior leaders. The skills that executives feared losing were already being carried out—often out of sight—by middle managers, whose value was being called into question. The researchers recommended that company leaders be more intentional and transparent about what they measure and recognize in the workplace. They also suggested involving more diverse voices in AI decision-making processes and planning carefully before adopting AI technologies. This includes deciding which parts of work should stay human, protecting time for learning from mistakes, using efficiency gains to give workers more time for meaningful tasks and better work-life balance, and maintaining high quality standards even as AI becomes more common.