If Facebook isn't developing a trust algorithm for news right now, it's a big missed opportunity.
Step 1: Crowdsource: Find an elegant way for users to report fake or misleading posts/pages on the Internet.
Step 2: Train the algorithm: Start with verified, authoritative accounts and then spider out to their high strength connections. Sites that get a lot of high confidence flags are themselves flagged as sources of misleading information.
Step 3: Apply results: Influence edgerank behind the scenes and/or be displayed as an annotation near the embedded preview.
"False story"
Google should do the same.
Better: They should both provide the data as an API for all apps and sites to use. Imagine a browser extension that surfaces this too. Chrome could surface it just like their malicious site warnings.
Ultimately this is similar to the email spam problem; although not exactly the same because there's even more incentive to game the system with false positives.