I need a lightweight proof-of-concept that shows how artificial intelligence can flag political misinformation on social media. The scope is intentionally lean: (even if work-30%/40% would be fine) doesn’t need to be 100% working • Platforms: focus only on public Facebook and Instagram content. • Targets: detect two threat types—false news articles shared in posts and misleading political images. • Outcome: the tool should output a simple confidence score and a short explanation (e.g., headline similarity, reverse-image findings) for each flagged item. A small, well-commented codebase in Python is ideal. You may combine off-the-shelf NLP models for text and a basic image-forensics approach (hash matching or lightweight CNN) for pictures. A command-line script or minimal web page that ingests sample URLs, processes them, and prints results is enough for this stage. Please include: 1. Read-me with setup steps. 2. Brief note describing the data or APIs used. 3. Suggestions on how this prototype could be scaled later (optional but appreciated). Keep it simple, runnable on an average laptop, and demonstrate the core logic clearly.