This year we saw the weaponization of our own data. Information is like uranium – it has huge potential value, but, as it decays, it becomes very dangerous. We have been hoarding all our data since the start of the digital records era over 30 years ago, and the decay is catching up to us.
We saw incredibly large spills this year, where decades worth of sensitive data was exfiltrated from large organisations and shared on the dark web. Ironically, these organisations hadn’t even really been mining that data and getting good value from it – they were mostly still holding onto it just because it was hard to get rid of. So, they were carrying all that risk, but not getting any value.
Since the breaches of Optus and Medibank in Australia, there is suddenly pressure on organisations to get rid of this uranium, before it can cause harm. And this is appropriate – we should not be keeping information that has risk, but no continuing value. But, we need to remember what happened to a generation of immigrants in the UK who were unlawfully stripped of statehood, and all the reports lost by the Japanese Supreme Court, including those of the 1997 Kobe child murders, because the records were destroyed too soon.
We all learned the hard way this year that keeping too much for too long is dangerous. But we need to remember lessons already learned about preservation, and make sure our priorities are balanced. This has previously been impossible to manage at scale, but is now easily achievable with AI. With Castlepoint we can now manage retention obligations, automatically find and dispose of risky data, and actually get value from what we keep. This means we enrich our uranium for our own benefit, and stop benefitting the bad actors.
This article by Rachael Greaves was originally published by Computer Weekly