1.“AI Slop”: Impact on Information Quality and Diversity
Artificial intelligence has made content creation faster, easier, and more cost-effective than ever before. Today, a single person can create hundreds of articles, images, or videos within a few hours and distribute them at virtually no cost. In 2026, Columbia University's Institute of Global Politics published the report 'AI Slop and the Information Ecosystem', which warns of the growing impact of ‘AI Slop’: huge volumes of AI-generated text, images, videos, and audio files that are produced rapidly, and serve primarily to generate clicks, views and advertising revenue rather than to inform, educate, or inspire, and are flooding digital channels on an unprecedented scale.
The challenge is not that AI creates content. The challenge lies in the fact that online platforms reward visibility, engagement, and reach. This creates incentives to produce as much content as possible in the shortest amount of time, rather than investing in originality, expertise, and quality. The result is a growing flood of content that often offers little that is new, yet still competes for the same attention as carefully researched articles, expert opinions, or creative original work.
This makes it more difficult for users to verify information and distinguish meaningful insights from generic, repetitive content. Over time, this can erode trust in digital content.
Ironically, AI systems were originally trained on content created by humans. Human creativity, experience, and expertise form the foundation upon which they are built. With the rise of synthetic content, however, AI models are beginning to learn from information generated by other AI systems. This creates a self-referential loop in which machines gradually recycle and recombine their own outputs. This is the paradox: AI learns from us, yet as human input diminishes, AI becomes repetitive, homogenized, and culturally stagnant, we ultimately starve the system of the very originality it depends on.
Ultimately, “AI Slop” can become a societal problem. When quantity becomes more important than quality, and when reach and ROI carry more weight than expertise and originality, society runs the risk of creating an information environment that reduces access to diverse perspectives, expertise, original thinking, and authentic human contributions, and weakens the quality of public discourse.
2. “Cognitive Offloading”: Loss of Learning and Thinking Skills
When AI takes over research, summarizes texts, or provides initial answers, it reduces the user’s own cognitive effort. At the same time, however, the need to analyze information independently, establish connections, or draw one’s own conclusions diminishes. The outsourcing of this mental work to AI systems is called “Cognitive Offloading.”
The ScienceDirect study “AI-overdependence and human cognitive decline: Hazards, evidence, and mitigation strategies” from 2026 examined the risks that can arise when people rely too heavily on AI tools and the extent to which cognitive abilities are impaired by excessive use. The study concludes that excessive dependence on AI can impair cognitive abilities such as critical thinking, creative thinking, decision-making, analytical thinking, memory, as well as attention and executive functions.
For example, people remember less because AI remembers for them, people analyze less because AI summarizes for them, and make fewer decisions because AI makes recommendations.
What impact does it have on a society when core cognitive skills are trained and developed less and less frequently? While today’s adults can still draw on skills they acquired long before the AI era, a generation is growing up for whom, for the first time, intelligent assistance systems are a given from the very beginning.
The EU Kids Online 2026 report examined the use of generative AI among more than 25,000 children and adolescents aged 9 to 16 in 20 European countries. The results show that about seven out of ten children already use some form of generative AI. Generative AI is primarily used for practical and educational purposes, to save time, get help with homework, and have complex concepts explained. Some children even stated that they consider AI-generated information to be more reliable than other online sources.
This development is also cause for concern at the political level. In its information report “Artificial Intelligence in Classrooms: Cognitive Dimensions,” the European Parliament warns of the so-called "AI learning paradox": while AI can improve the quality of work outcomes, actual learning declines at the same time. Not only are basic skills such as reading, writing, arithmetic, and problem-solving practiced less frequently as a result, but attention, critical thinking, self-regulation, and the ability to make independent judgments could also be weakened in the long term. If cognitive tasks are permanently delegated to AI during a student’s school years, there is a risk that these skills may not be fully developed. Therefore, the report emphasizes that AI in education must be carefully guided and should not take over tasks unchecked, but rather should provide targeted support for learning processes.
The authors of the ScienceDirect study reach the same conclusion: AI should serve only as a tool, not as a replacement, and should be viewed solely as a supplement to traditional learning methods. In that case, it can support learning and thinking processes and promote the development of important cognitive skills.
3. “Empathy Gap”: Effects on Social and Emotional Development
AI systems are increasingly becoming conversation partners, advisors, learning guides, or even digital companions in everyday life. Their responses often come across as understanding, supportive, and empathetic. In the EU Kids Online study, a small but significant proportion of children report using chatbots as conversation partners or for advice, especially when they are bored or lonely. But how does it affect human development when AI replaces genuine human interaction?
The 2026 essay “Complexities, Gaps, and Risks in Simulating AI Empathy” by Springer Nature points out that, according to scientists, AI can simulate certain aspects of empathy, particularly what we call cognitive empathy, that is, the rational understanding of other people’s thoughts, perspectives, or feelings without actually sharing those emotions. However, its ability to simulate authentic empathy, that is, the genuine, unsolicited feeling and understanding of another person’s emotional world, is limited. In this context, the term “Empathy Gap” is used. For example, the authors warn that children who spend a disproportionate amount of time with AI agents rather than with other people may be at risk of impaired development of empathy and prosocial behavior. This is because genuine human empathy is not an innate, fully formed ability, but rather a complex skill that develops throughout childhood and adolescence through social interactions, shared experiences, and the formation of interpersonal relationships.
If these experiences are lacking over an extended period, important social skills such as perspective-taking, compassion, and interpersonal understanding may develop to a lesser extent. On a societal level, this could also have implications for social cohesion, cooperation, harmony, and peace.
