I was about to say that I find this article on Reuters quite surprising. It suggests that many large companies employ people to read outgoing email to ensure that there is nothing that could harm their reputation. It’s not actually surprising and has been happening in most heavily regulated companies for years.
I think it’s just the sensationalism, jumping on the back of the growing wave of privacy concerns. It’s easy to get caught up, but the idea that an organisation could employ a sufficient number of people to filter a significant proportion of email is absurd. However, organisations have been filtering email and reviewing potentially harmful content for many years. This does not mean that they are “reading” everything.
I was putting in email filtering systems for a company I worked at nearly 10 years ago. The filter used a system of points for key words. If an email went above the trip figure (e.g. 3 keywords totaling 10 points) then you could quarantine it for checking, or if it was so obviously dangerous then you could bounce it back to the sender with an appropriate message.
This is a necessary part of managing internet services for a corporate entity. For example, a company advisor may suggest the projected rate of return on a particular investment but they should not guarantee a return. If an advisor were to send a message to a client that included the word guarantee then it may be quarantined for review or returned to the sender to amend. Not doing this could mean that the company is not taking appropriate precautions to protect it customers, employees (from mistakes) and it’s reputation.
This might sound like big brother type snooping, but a corporate email service is owned by that corporate body and it has a right to scan data transmitted through it. Many companies allow some level of personal use but will disclose in their usage policy that the content is scanned.
Finding that Internet Service Providers are doing this for consumer services would be of greater concern. You are not sending email in their name, and many will have separate domain names (.net and .com) for employees and public services.
Filtering of email by consumer ISPs may also imply a responsibility for the content of that email. If you read something and let it through, are you approving it? I wonder if we will see a test case for Googlemail and their context sensitive advertising or if they have resolved this already.