Travis,
Its all lawyer speak. The lawyers require companies put that kind of thing in to CYA and because it keeps them employed. Also like any powerful tool it can be used wrongly. If I sold chainsaws I would have to include in my terms that I don't guarantee the chainsaw will not slip and cut off your leg, not because I would not design every safety feature I could into my chainsaw but because I know someone out there will find a way to disable my safety features and cut off their leg and then they will blame me.
From the main S3 page:
Data stored in Amazon S3 is secure by default; only bucket and object owners have access to the Amazon S3 resources they create. Amazon S3 supports multiple access control mechanisms, as well as encryption for both secure transit and secure storage on disk. With Amazon S3’s data protection features, you can protect your data from both logical and physical failures, guarding against data loss from unintended user actions, application errors, and infrastructure failures. For customers who must comply with regulatory standards such as PCI and HIPAA, Amazon S3’s data protection features can be used as part of an overall strategy to achieve compliance. The various data security and reliability features offered by Amazon S3 are described in detail below.
Also from this page and more what I was talking about:
Amazon S3 provides a highly durable storage infrastructure designed for mission-critical and primary data storage. Objects are redundantly stored on multiple devices across multiple facilities in an Amazon S3 Region. To help ensure durability, Amazon S3 PUT and COPY operations synchronously store your data across multiple facilities before returning SUCCESS. Once stored, Amazon S3 maintains the durability of your objects by quickly detecting and repairing any lost redundancy. Amazon S3 also regularly verifies the integrity of data stored using checksums. If corruption is detected, it is repaired using redundant data. In addition, Amazon S3 calculates checksums on all network traffic to detect corruption of data packets when storing or retrieving data.
I know my local data is not stored in multiple locations instantly. We upload database backups to S3 overnight and push hourly backups to a different server from the database server, but this is a far cry from the kind of redundancy amazon provides.
We as a company happen to like controlling our own servers. It works for us and makes economic sense, and we really don't NEED the kind of redundancy Amazon offers, but that does not mean Amazon servers can't be trusted.
Though I will grant that if I were to move my servers to amazon I would do the opposite of what I do now, and push my backups down from amazon to a local backup devise every night, but that's just the DBA in my and DBA's are paid to be paranoid.