Readers can't distinguish human vs AI creations, undermining trust in blogs and news. A verification tool could analyze writing patterns and image artifacts to estimate likelihood of AI generation.
Build a lightweight browser plugin that highlights suspected AI content with confidence scores. Include explanations of detection criteria to educate users.
Publishers and educators would pay for API access to vet submissions. Individuals might tip for one-time checks.
Start with text analysis focusing on common LLM patterns before tackling images.
Risk is an arms race as AI improves at mimicking humans, making detection harder.