An experiment involving four commercial chatbots across Europe found that these often blur the line between trusted health information and advocacy, frequently directing users to anti-abortion content without clearly identifying their sources or distinguishing official authorities from others.