Opt-in system to filter offensive email content

December 6, 2021

To minimize the risk of emails containing hate speech, threats, and other harmful content reaching members of our community, Harvard University Information Technology (HUIT) has partnered with the Office of Equity, Diversity, Inclusion, and Belonging to offer an opt-in system that detects and adds a warning label to incoming emails containing potentially offensive content.

The system detects words in incoming emails that are commonly used in abusive or threatening ways to demean, harass, or express prejudice against individuals or groups, including racial, gender, religious, and sexual slurs. When triggered, a “[Potentially Offensive Content]” warning label will be added to the subject line of the email before it is delivered to the intended recipient. You can also set up an inbox rule to direct flagged messages to a separate folder. Learn more about the system and how to opt in.