With the spread of false and misleading information in current news, many\nalgorithmic tools have been introduced with the aim of assessing bias and\nreliability in written content. However, there has been little work exploring\nhow effective these tools are at changing human perceptions of content. To this\nend, we conduct a study with 654 participants to understand if algorithmic\nassistance improves the accuracy of reliability and bias perceptions, and\nwhether there is a difference in the effectiveness of the AI assistance for\ndifferent types of news consumers. We find that AI assistance with\nfeature-based explanations improves the accuracy of news perceptions. However,\nsome consumers are helped more than others. Specifically, we find that\nparticipants who read and share news often on social media are worse at\nrecognizing bias and reliability issues in news articles than those who do not,\nwhile frequent news readers and those familiar with politics perform much\nbetter. We discuss these differences and their implication to offer insights\nfor future research.\n