Cloud-based computing models make a great deal of sense for plenty of businesses -- especially those small-to-medium-size ones that don't want to sink a fortune into building and maintaining an IT infrastructure. By leveraging economies of scale, cloud storage can be cost-effective, letting a business pay only for the computing services it needs, as it's needed.
But using the cloud model means losing a degree of control over the information that's stored. That isn't necessarily a bad thing; the cloud service provider might manage the information better than its customer would or could. But loss of control can lead to big problems when a business finds itself in litigation.
Whenever a business or individual reasonably anticipates litigation, it has a duty to preserve any information relevant to that litigation. While this may sound like legal mumbo-jumbo, in practice it means that whenever a business is confronted with a situation where it finds itself worrying about a lawsuit, it needs to act swiftly to make sure any information related to that potential lawsuit is preserved. This can be tough to do when your information is poorly managed -- and the legal consequences for failing to preserve information are severe.
When dealing with tangible objects, the duty to preserve is relatively easy to handle. For example, if a guest slips and falls on a rug in a hotel and is injured, the hotel has a duty to preserve the rug somewhere safe. Since the hotel can reasonably expect the customer to sue and argue that the rug was somehow unsafe, if the hotel destroyed or lost the rug after the guest fell, the guest would not have access to the evidence the guest needed to prove that something was wrong with the rug that caused the guest to slip. Thus, the hotel cannot just destroy the rug -- and if the hotel were to do so, a court could sanction the hotel (meaning that the guest could potentially win the argument that the rug was unsafe simply because the hotel threw it out).