I am envisioning a chat session where my little sister is chatting away with someone she met online and that person begins to make sexual advances. Instead of her having to take time to get offline, remember everything that was said during the conversation and only armed with that person's screen name and testimony of the conversation that took place, she could simply click a button on that chat session window which would pull in the aim user handle of both parties, the contents of the conversation window, as well as the server which initiated the chat and the remote ip of both parties involved, do a whois on the isp of those ip addresses, and automaticly post that information to a online form such as the one found at https://secure.missingkids.com/missingkids/servlet/CybertipServlet?LanguageCountry=en_US.

I realize there may implications to doing something like this such as abuse by anyone having a particular bias against another but with the ability (according the aim sdk) to read the contents of chat session window without user interaction, I can forsee a system where the conversation in question can be processed and assigned a severity level to help drive traffic down, I can also see a disclaimer for the user that is flagging the conversation to have to agree that the concern is a valid one.