It has now emerged that online accommodation and activities marketplace Airbnb has filed a patent with the European Patent Office (EPO) for a new technology that will scour the internet - including social media accounts - to calculate the risk of someone trashing a host's home.
The 'trait analyser' software will use artificial intelligence to mark down people 'associated' with drugs or alcohol, hate websites, or sex work by scanning keywords, images and video footage across the internet linked to a potential customer to assess their trustworthiness.
The technology is designed to scan the online profiles of would-be bookers to judge whether they will be reliable customers or not, according to EPO documents. The most recent patent is dated 2019, but initial plans were first put forward in 2014 and are thought to be initially linked to the Trooly start-up that was acquired by Airbnb back in 2017.
Airbnb already makes a risk assessment on every one of its reservations. Its website explains: “We use predictive analytics and machine learning to instantly evaluate hundreds of signals that help us flag and investigate suspicious activity before it happens.”
As a customer we accept that some form of assessment is being made – that is standard safety for everyone. However, this ‘trait analyser’ may appear to many to be a more intrusive. The programme will apparently assess guests’ 'behavioural and personality traits' including 'conscientiousness and openness' complementing the platform’s existing credit and identity checks.
Negative traits the software will be on the lookout for include 'neuroticism and involvement in crimes' and 'narcissism, machiavellianism or psychopathy' as they are 'perceived as untrustworthy', the company has said.
Among the software's capabilities will be the ability scan news stories that mention a particular name, which could potentially score you negatively for another person’s crimes or misdemeanours. Airbnb says all the data will be used to predict how the customer will act offline and will be cross-referenced with information including 'social connections', employment and education history. This, it hopes, will all it to better calculate the compatibility of host and guest.