How do I keep AI-written social media posts from looking like AI or getting flagged?
Whether a platform such as Meta treats a post as AI content depends on the content, since each platform has its own algorithm. Problems with content on social media usually come from the algorithm not liking it or from people reacting badly to obvious AI writing. For example, em dashes make it obvious that AI wrote a post, since there is no em dash key on a keyboard.
What helps:
- Tell the AI assistant to rewrite the post using the humanizer skill. It is built into the assistant and removes signs of AI generated writing, such as em dashes.
- Post quality content. Have the AI make several ideas or images and pick the best ones instead of posting everything.
- Review before posting. Tell the assistant to create the posts as drafts on your social media calendar, for example 30 days of content, then approve the ones you like and delete the others. You can fully automate posting, but there is a risk it posts something you do not like, so human review is a good idea.