AI can summarize meetings, but here’s what it still can’t do in 2026 (www.techradar.com)

In a recent analysis, experts highlighted the ongoing limitations of AI, particularly in the context of large language models (LLMs) like ChatGPT and Claude. As AI tools are increasingly leveraged for tasks such as summarizing meetings and generating content, it’s crucial to understand what they still struggle to achieve as of 2026. Key shortcomings include their inability to acknowledge uncertainty or admit ignorance—often leading to “hallucinations,” where they confidently present false information. This tendency stems from their design, where they predict language rather than retrieve factual data, underscoring the importance of critical fact-checking, especially in high-stakes scenarios such as legal or medical consultations. Moreover, AI remains inadequate in roles that require understanding human experience, such as therapy and moral reasoning, as it lacks consciousness and lived experiences. Its failure to process real-time updates also raises concerns, particularly for applications in journalism and fast-moving fields, where outdated information can be presented with misplaced confidence. Recognizing these limitations is essential for users, enabling them to employ AI tools more effectively and cautiously while pursuing their goals. This critical perspective is necessary not just for better deployment but also for fostering a clearer dialogue about the potential and responsibility associated with AI technologies.