How about not actively providing false information, but being allowed to respond with wrong information?