What are the ethical questions posed by insurers' potential use of social media to gain more information on customers? Ralph Savage talks to Post Online.
Sir Tim Berners-Lee told insurers at the British Insurance Brokers' Association conference in May to think before they use social media data to help them underwrite policies and manage claims.
As someone who believes in the individual rights and freedoms of citizens I instinctively agreed with Sir Tim's note of caution. However, it appears there are those who don't.
I pointed out at a recent Post and Chartered Insurance Institute roundtable discussion how permission to use data would become a significant area of debate for insurers and brokers given the prevalence of publicly available information at our fingertips.
Gravytrain's Kevin Taylor therefore seemed to be asking a pertinent question in his subsequent article Facebook is your Friend in which the question apparently was whether insurers could, or should, use data gathered from social networks to compile quotes.
He said pointedly: "Insurers must remember one thing: there is no issue with monitoring data consumers have willingly put into the public domain."
Taylor discussed the possibilities that social media data allows insurers to decode our language, the reliability and trustworthiness of our real-life social networks and information about our lifestyles, all through online social media monitoring that he believes would go a long way towards pricing risk accurately for many types of insurance policy.
Nevertheless, while the ‘could' is very well catered for in his argument, little of the ‘should' is contained. Perhaps that is why Sir Tim's comments at Biba need to be juxtaposed with any argument advocating a headlong rush into applying algorithms that supposedly deliver an "immediate response on an individual's risk level", as Taylor suggests.
The inventor of the World Wide Web sits on the Transparency Board of Open Government at the government's data website and he is a staunch advocate of opening up publicly owned data, but he believes the private sector has a major responsibility in how it uses the information available to it.
He said: "Insurance companies should come up with guidelines to reassure people about what information will be used.
"For example the industry could say: ‘We will not use the information you share on social networks from when you are under 18. We would like this to be the case, as childhood should be protected.'"
At the heart of Sir Tim's message was the issue of data provenance. He emphasised that users of data like insurance companies had an obligation to acknowledge where the information they are planning to use is coming from and how the data should be used when they get it.
"Companies that deal with a lot of data need to track where the data came from; the appropriate use and destiny of that data is important," he said.
It doesn't take a genius to realise that the provenance of data on social networks is not always clear, nor is its accuracy. Given that fact, who is to judge what ‘trustworthy' or ‘reliable' means in the context of an algorithm applied to your network of friends when such concepts are purely relative?
This issue could be of even greater significance for insurers as next year sees the introduction of a new regime around disclosure at point of sale and the removal of the duty of good faith from consumer insurance contracts.
The Consumer Insurance (Disclosure and Representation) Act 2012 will permit ‘honest misrepresentations' to be made and, conceivably for those in the camp that believe social media data is fair game, this could be an opportunity to dispute claimants' sincerity.
I understand more than most that it can be positive for insurers to have access to data that helps them price risk effectively or limit the effects of fraudulent claims.
I also understand that we are in a new communications age where publication of data about ourselves has raced ahead much faster than the general population's understanding of the risks involved.
It therefore seems exploitative for insurers to take advantage of that naiveté. There must be limits to the way our publicly available data is accessed and used by private organisations such as insurance companies.
Moratoria have been put in place for sensitive data issues like this before and the industry has proved itself to be sensitive with regards to issues like genetic records. I hope it can also be on this topic too.
I am as enthused about social media as anyone, but a free for all backed by the resources of a £20bn industry should not be enabled without a substantial ethical debate.
Ralph Savage, director, RTS Media
- ABI previously warned government over hazards of flammable cladding
- Towergate group company to launch under new brand name
- Insurers issue warning as government to test cladding on 600 buildings
- Data analytics will impact property underwriting but could lead to 'uninsurable' properties
- Full report: Motor Report: What's driving motor?
- Marsh Networks’ David Hopwood on the seven issues a reduced discount rate creates for SME brokers
- Government resurrects whiplash reform