AI Agents and Cultural Blind Spots: Culture as Security Infrastructure
AI agents do not only fail because prompts are vague. They can also fail when culture, language and context are treated as universal. These blind spots can turn small misunderstandings into governance, safety and security risks.