The experiment, called TwiGUARD, lets Twitter users check if a follower is a spammer or if a link embedded in a tweet is malicious.
From the article:
TwiGUARD also tracks the Twitter threat level, which will be adjusted according to active attacks on the service, providing attack details and how users can protect themselves. The threat level starts with Threat Level 1 (normal activity), then Threat Level 2 (possible threat), such as a jump in malicious activity or an indication of a larger attack in the works, and Threat Level 3, in which specific attacks are found on Twitter.
The TwiGUARD site, which is still under development, uses a reputation-based algorithm to see if a follower is a spammer or if a link is infected.
More information in this blog post by Errata’s David Maynor.