ARTICLE · 1105799
AI普及趋势下的心理健康问题挑战 Challenge of mental health amidst popularization
AI普及趋势下的心理健康问题挑战 Challenge of mental health amidst popularization心理健康，正在成为越来越多国人无法回避的话题。在一个竞争激烈、同时经历了深刻经济与社会变迁的社会里，许多人都受到心理健康问题的困扰。而AI普及率的提高，可能会加剧社会的心理问题。 Mental health is becoming an increasingly unavoidable topic for many people. In a highly competitive society that has undergone deep economic and social change, many people struggle with mental health. The increasing penetration of AI may exacerbate the emergence of psychosocial problems in society furthermore. 一项发表在英国医学期刊《柳叶刀》上的研究估算：2023年，中国约有1.91亿例精神障碍病例，相当于大约每7个中国人中就有1个。与1990年相比，这一数字增加了46%，剔除人口老龄化以及总人口规模变化的影响后，估算患病率仍然上升了14.6%。由于这些数字是基于相当有限的数据和诊断结果建模估算的，所以很可能低估了真实情况。 A study published in the Lancet, a British medical journal estimates mental disorders’ prevalence in China, with 191m cases in 2023 (equivalent to about one in seven Chinese). That was 46% more than in 1990, but after allowing for the country being older and more populous, the estimated prevalence was 14.6% higher. Because these are modelled estimates, based on sometimes sparse data and diagnoses, these figures are likely to be an underestimate of the true picture. 最近，美国多所知名大学的研究人员在《美国医学会杂志·网络开放》期刊上发布的一项2万人参与的研究显示，AI使用频率与抑郁表现程度具有相关性，每日使用AI的用户患中度及以上抑郁症的概率要高出30%。对原本已经很严峻的社会性心理健康问题，这无异于雪上加霜。 Recently, researchers from several well-known universities in the US published a study involving 20,000 participants in the Journal of the American Medical Association - Network Open, revealing a correlation between the frequency of AI use and the degree of depressive symptoms, with daily AI users having a 30% higher probability of suffering from moderate or severe depression. This is akin to adding insult to injury to the already severe social mental health issues. 中国青少年研究中心2025年对全国7个省份的8500余名未成年人进行了调查，结果显示，有六成以上未成年人使用过AI。调查中，21.5%的孩子表示“我只想和AI聊天，不想和真人聊天”，还有近半数孩子心中有烦恼时，只愿意和AI说。 The China Youth and Children Research Center conducted a survey in 2025, involving over 8,500 minors from seven provinces across the country. The results showed that over 60% of the minors had used AI. In the survey, 21.5% of the children stated, "I only want to chat with AI, not real people," and nearly half of the children were only willing to talk to AI when they were troubled. 青少年过度与AI交流，可能削弱现实社交能力： 与AI交流缺少肢体语言和真实情绪，AI完全顺着用户的意思回应，让人感觉舒服，但长期沉浸其中，可能会失去与人交流的能力。 此外，AI会使用一些心理咨询技术，如先理解、共情，再顺着用户的话说，但它不能识别危险，也不能身临其境，无法提出真正个性化的建议。有一定识别能力的人可以把它当作辅助工具，但缺乏判断力的人容易全盘相信AI，甚至依赖成瘾。 Excessive interaction with AI among teenagers may weaken their real-life social skills: AI communication lacks body language and genuine emotions, and responds entirely according to the user's intention, making people feel comfortable; but if they immerse themselves in this for a long time, they may lose the ability to communicate with others. In addition, AI employs some psychological counseling techniques, such as understanding and empathy, and then echoing the user's words. However, it cannot identify danger, nor can it immerse itself in the situation, making it unable to provide truly personalized advice. People with certain recognition abilities can use it as an auxiliary tool, but those lacking judgment are prone to fully trusting AI and even becoming addicted to it. 成人使用AI虽然更成熟、多样，但也一样需警惕AI成瘾：医疗临床不乏因焦虑而反复查询AI、越查越焦虑的案例，甚至发展为不相信医生、不听遗嘱；另一种则是事事依赖AI，懒得自己思考。如果一个人不再自己思考，甚至对结果无所谓，背后的原因可能是本身存在焦虑或抑郁情绪。 Although adults are more mature and diverse in their use of AI, they still need to be wary of AI addiction: there are many cases in clinics where anxiety leads to repeated AI searches, which only exacerbates anxiety. Some patients even refuse to believe doctors and argue with them. Another situation is relying on AI for everything and being too lazy to think for oneself. If a person no longer thinks for themselves and is even indifferent to the results, the underlying reason may be anxiety or depression. 中国精神疾病的识别率和治疗率一直非常低。2021年发表在《柳叶刀》上的一篇论文发现，在接受调查的1,007名患抑郁障碍的中国人中，只有9.5%接受过治疗，而得到充分治疗的只有0.5%。除此以外，服务能力、费用负担以及有效监管等问题仍然存在。按人口比例计算，中国的心理治疗师远少于西方国家。 The rates of identifying and treating mental illness keep very low. Another paper in the Lancet, in 2021, found that of 1,007 Chinese surveyed with a depressive disorder, just 9.5% got treatment, while a mere 0.5% were treated adequately. Yet problems of capacity, affordability and sound regulation remain. China has far fewer pyschotherapists per head of population than the West. 面临这些严峻挑战，希望在哪里呢？ Facing these severe challenges, where is the hope? 首先，政府计划到2030年建立覆盖全国的心理支持和危机干预体系，让更多人能够获得药物治疗和心理治疗。相关计划包括：每所学校都设立心理咨询室、每个县至少有一家医院提供精神科服务，以及培养更多心理健康专业人员。 First，the government now aims to build a nationwide system of psychological support and crisis intervention by 2030, giving more people access to drug-based treatments and psychotherapy. Plans include counselling rooms in every school, psychiatric services in at least one hospital in every county and more mental-health professionals. 其次，公众意识有所提高，尽管长期以来，社会污名以及不愿承认精神疾病的态度一直普遍存在，但寻求帮助的人正在增加。 Second, greater public awareness can be observed, though social stigma and a reluctance to acknowledge mental illness have long prevailed, calls for help are growing. 再次，利用AI技术，将AI技术运用在早期识别阶段，并最终建设成为专业技术主导的、医疗和软件共同搭建的正规平台。目前已有线上认知行为治疗系统引入AI技术，可以跟用户聊天、布置任务，捕捉消极言语，用于监测儿童青少年的心理状态。有些项目还提供类似“树洞”的工具，学生可以倾诉，软件会自动监测敏感字眼，并在后台报告，同时配备数字心理老师。 Then，utilizing AI technology, we can apply it in the early identification stage and ultimately build a professional, technical, and formal platform jointly established by medical and software industries. Currently, online cognitive behavioral therapy systems have incorporated AI technology, which can chat with users, assign tasks, and even capture negative language to monitor the psychological state of school children and adolescents. Some projects provide tools similar to "tree holes", where students can go to talk, and the software will automatically monitor sensitive words and report them in the background, while also providing digital psychological teachers. 最后也是最重要的，公众需要客观看待AI，把它当作工具，而不是一味依赖；在使用AI的基础上核实信息、验证知识，而不是被工具牵着鼻子走。 Lastly and most importantly, the public needs to view AI objectively, treating it as a tool rather than relying on it blindly; they should verify information and knowledge when using AI, rather than being led by the tool.