We harnessed Facebook’s new facial recognition and auto tagging technology to search for missing persons. By creating Facebook profiles for the missing and building them a friend network across the world, we created a social search party that scoured the backgrounds of millions of photos and videos posted on Facebook, daily. If a missing person were to appear in the background of an image uploaded to Facebook by anyone in their friend network, the facial recognition tech would identify them and MPAN would be notified. Followed by any appropriate investigation.

Simply put, we asked the world to donate their Facebook friendship to enable advanced AI to search for missing persons.

To harness Facebook’s facial recognition technology, we created Facebook profiles for missing people and populated them with personal information and photos of each individual. Tagging each photo trained Facebook on what face to look for. A landing page invisiblefriends.com.au was established to explain how the tech and the initiative work, while also providing authenticity to the profiles.

Then a large social and PR push aimed to garner as many Facebook friends as possible for each Invisible Friend profile, from every corner of the globe.

The campaign launched in mid April and is ongoing, with missing persons organisations around the world wanting to implement the program to help bring home some of their missing people.

The campaign reached two thirds of the Australian population and gained media coverage in more than 30 countries around the world.

At the time of this submission, (only two weeks after launch) each Invisible Friends Facebook profile had more than triple the average amount of Facebook Friends and climbing. With a joined total of over 10,000 friends, searching through tens of millions of photos and videos posted by friends and friends of friends, each day.

Invisible Friends is being rolled out by other missing persons organisations in the USA, Britain, Europe and Asia.

Most impressively, on the first day of the PR launch, based on the overwhelmingly positive public and media response, Facebook's Communications Department reached out to MPAN offering praise and help for the international roll out. Negotiations are underway.

We’re more connected than ever before, thanks to Facebook and social media; yet the way we search for missing persons hasn’t changed.

We rely on people to ‘look for’ or help ‘spot’ missing persons in public; which is why we have always defaulted to posters, milk-cartons and other low-cost, high awareness media.

Our problem wasn’t with the media, but the requirement of people to ‘look’ for missing people - the reliance on others to find the proverbial needle in a haystack.

In an ‘always on’ world, we’re bombarded with 1000s of messages daily, and the chances of achieving cut-through is diminishing. We utilised technology that resulted in machines doing the searching for us.

Using Facebook’s facial recognition - an AI engine that analyses every friend’s photo and video for your face - We made the process of search for ‘missing persons’ as simple as adding a friend on Facebook.

Over 38,000 Australians go missing every year. The impact of these events are profound with recent research from Missing Persons Australia showing that for every one person that goes missing, 12 others are directly affected - well over 500,000 people each year.

After the first 48hrs, the chances of finding these missing persons diminishes significantly. Evidence and the public consciousness dries up and police are left with cold leads and nothing new to investigate. So they move on to other cases.

Our brief from Missing Persons Advocacy Network, who support those left behind and work to generate awareness of the people who are missing long-term, was to find a way to generate new information, clues and awareness in these cold cases. To do so, would provide hope and support to the families of the missing and raise awareness of each missing person’s circumstances around their disappearance.