Facebook's army of chatbots could soon start messaging users based solely on which websites they have previously visited.
The social network's Bots on Messenger platform, which was announced during its F8 developer conference earlier this month, is currently in testing mode but there has been confusion around whether or not businesses will be able to send chats to people they haven't previously been in contact with.
Brands such as KLM airlines have been among the first to sample the bots, which let users start a conversation with brands to receive automated customer support, e-commerce guidance and more.
Upon the launch, Facebook said companies would be charged to send messages to people who had already voluntarily entered a dialogue with them, and that the tool was being heavily moderated to avoid users falling victim to spam.
However, some users who have been selected to sample the features have reported receiving sponsored posts out of the blue, based solely on their browsing history.
Facebook user and PR Katelin Schwarck wrote in a blog that she was surprised when she got a message request (below) from a name she didn't recognise, only to discover it was an ad.
The user hadn't 'liked' the Tiek's page on the platform, but said she has previously visited the firm's website; showing that Facebook is toying with serving ads to users in Messenger depending on which sites they've been looking at.
The social giant has been keen to reassure Messenger's 800 million users that it will head off the obvious spam potential of such a system.
"We are testing sponsored messages from a small number of businesses," a Facebook spokesperson said in a statement, adding: "People will always have the ability to receive or block messages from each business, and we will be reviewing feedback to businesses as these tests progress."