ChatGPT's vulnerability was exposed when a simple optical illusion tricked it into erroneous responses, highlighting broader AI limitations like hallucinations and lack of true comprehension. This raises concerns for reliability in critical sectors, urging hybrid human-AI systems to balance innovation with accountability.
