The question of who owns the data is partly determined by the kind of cloud infrastructure that is in place for your business.

The advantages that cloud based hosting brings needs to reiteration. Cloud hosting is infinitely scalable, needs radically lower CapEx and with distributed hosting, it is much more fail-proof than data hosted on your premise. But the recent legal spar between Apple and the US government has brought the limelight once again back on the debate about ownership of data. In the case of Apple, the fight is over decrypting the iPhone used by a criminal in order for the government to hack in. But the case may as well be between law enforcement and a web host over data stored by your customers on your startup service.

In such a case, who owns the data? Before we discuss this, it is important to identify the various kinds of cloud hosting services presently available. The first and the most popular among start-ups is the public cloud. In this instance, start-ups host data over virtualized servers located across the world. This is best suited for web services that anticipate server crashes due to peak volume loads. Since the cloud is distributed, there is very little chance for websites to crash due to a sudden influx of visitors. The second is a private cloud where you own or rent the infrastructure from a third party provider, but the data itself belongs to yours and no third party has access to it. The final alternative is the hybrid cloud which is essentially a mix of public and private clouds - part of the data is located on your own server while the rest is on public clouds.

As you may have guessed, the question of who owns the data is partly determined by the kind of cloud infrastructure that is in place for your business. But even before we try to tackle this question, the first thing to look at is who creates the data. If your customer data includes information patently created by the customer (like photos, videos, text documents, etc.), then the original creators of these documents continue to own copyrights of such data. However, this does not automatically mean protection of this data which continues to be determined by where it is hosted.

In the case of a private cloud server, the web host provides with only the infrastructure and the company renting the cloud owns all the data that resides in the server. However, this line begins to get fuzzy in the case of public and hybrid clouds. As a matter of fact, one of the major reasons hybrid cloud servers exist is to comply with local government regulations on data ownership. For instance, if you operate a business in Indonesia and use Microsoft Azure for hosting data, then it is likely that the data itself is hosted in a country like Singapore. In order to meet local compliance, you may be required to host a private server in Indonesia that stores a part of the customer database locally in your private server. This way, you own complete ownership of the data in your private server which local governments may demand access to depending on local laws.

In terms of data ownership, public clouds have the least confidentiality. Depending on the service you sign up to, the web host may own rights to at least part of your data. Even if your startup does legally own the data hosted on your behalf, a lot of providers still hold rights to what are called "emergent data" - that is, the manipulation of data hosted by the customer into distinct datasets that may be used by the provider for their own marketing or business purposes.

Cloud computing and storage is still an evolving segment and the laws surrounding this are not as robust as you may want them to be. While it may take years before sophisticated regulations are built around data ownership, it is always a good idea for start-ups and businesses to consult their lawyer before signing the terms of service of independent third party providers.

• SQL Data Services. Your Database in the Cloud.
• Encrypted Database in the Cloud: Practical Considerations
• Danger in the Clouds