Summary Despite the billions of dollars invested in digital transformation, most AI customer service applications continue to overlook crucial human factors essential for effective service delivery. This issue came to light during the author's experience in 2013 while leading a team to create the Universal Credit digital service for the British government. The project aimed to simplify a complex welfare system, making it user-friendly so that applicants could easily access help without calling hotlines or visiting employment centers in person. However, the rapid project planning and journey mapping processes neglected a fundamental fact: applying for welfare often occurs during pivotal moments in people’s lives, involving high emotional complexity due to issues like relationship breakdowns, illness, unemployment, or the death of a loved one. By designing the service to minimize human-to-human interaction, the team failed to address the emotional needs present in these critical situations. High-quality research has documented similar failures in digital transformation efforts, yet these findings have often been ignored. Over the past few decades, early chatbots and automated services promised a human-centric design but have not widely delivered on this promise. Organizations still aim to reduce labor costs while improving customer service efficiency, but this approach can backfire by undermining the social function of customer service. Customer service is more than just transactional; it involves social connections and emotional bonds that often become strategic or structural elements in business operations. When these human elements are stripped away to streamline processes, the effectiveness of the service diminishes. AI customer service implementations frequently encounter issues in several specific areas: Disruption of Social Function: Design flaws make it harder for people to get the help they need during emotional distress. Simulation of Humanity: Pursuing convincing human-like interfaces can lead to inauthentic interactions. Exploitation of Vulnerable Groups: Users with lower digital skills, language barriers, or who are in vulnerable positions find AI services particularly challenging. Erosion of Social Skills: Replacing genuine human interaction with purely transactional communication can diminish our ability to form meaningful connections. Based on the author’s extensive experience in both governmental and private sectors, effective AI customer service is not about replacing humans with machines but integrating their strengths. AI can offer consistent and round-the-clock service, handling simple tasks efficiently, while human representatives can manage complex situations, showing genuine empathy and addressing nuanced social issues. Smooth transition protocols ensure a seamless switch between machine and human modes, enhancing the overall customer experience. The future trend should move away from AI mimicking human interactions and towards a collaborative model where AI and humans work together based on their respective strengths. Organizations that recognize this will likely succeed better in their customer service endeavors. After decades of experimentation, the author emphasizes that technology works best when it complements rather than replaces human qualities. An example from his work on designing trauma-informed police reporting systems highlights the importance of a balanced approach, where AI and humans collaborate to support and enrich user experiences. In the realm of grief support, recent developments have sparked debate among experts about the role of AI. A seasoned grief counselor observed an increasing number of bereaved individuals turning to AI chatbots and digital companions for solace. AI tools offer benefits such as being available at any time, reducing initial barriers to seeking support, and providing consistent information. However, counselors are concerned that people may mistake algorithmic responses for genuine human listening, which is crucial in grief management. Grief is not just about emotional regulation but also about companionship and presence. Despite technological advancements, AI struggles to understand and respond to deep emotional needs, often providing superficial and inauthentic responses. Through collaborations with Cancer Research UK, researchers found that users can instinctively differentiate between machine-generated and human responses, a skill vital for true emotional resonance. Another significant issue is that AI cannot fully comprehend the complexity and depth of grief. Misleading or inaccurate advice can worsen the distress of those already suffering. AI lacks the ability to witness and empathize with another person's pain, which is a critical aspect of healing. Privacy concerns are also paramount, as people in deep grief may inadvertently share sensitive data without fully understanding its implications. This data can be used to train commercial systems, raising ethical issues about consent and data use. Experts recommend that AI grief support tools should be designed to guide users towards real human connections, not replace them entirely. These tools should transparently acknowledge their limitations and direct users to human experts when appropriate. Furthermore, AI should assist human professionals by reducing administrative burdens, enabling them to focus more on providing empathetic and meaningful support. Developing AI for grief support requires extensive input from grief experts and actual cases of loss to ensure the technology captures the nuances of the grieving process. Data collection from bereaved individuals should be conducted with strict ethical guidelines, given their decreased decision-making capacities during emotional lows and vulnerability to exploitation. In conclusion, an experienced grief content expert warns against prioritizing technological advancement over genuine human connection in grief support. While AI has value and specific applications, it cannot replace the empathy and support that only humans can provide. The expert’s background includes collaborations with multiple charitable organizations and law enforcement agencies to create services that truly support users emotionally.