At Visions Federal Credit Union, VP and CIO Tom Hull decided to turn over 24/7 security monitoring to Perimeter eSecurity, but still keep the company’s own firewalls. “I think there is a hard sell there,” says Hull, whose Endicott, N.Y.-based company has just 400 employees and annual sales of $80 million, of the in-the-cloud model. “We still retain their help in managing the firewalls, but we didn’t want to rely on the schedule of a third party to institute any changes in our environment. Plus, as it relates to any outages, downtime, system maintenance or things of that matter, that was another thing we could not relinquish control of.”

In London, AT&T customer Martin Joy also decided against AT&T’s virtual devices. “I’m not keen to see a device on my premises. The important thing is to make sure the technology makes sense and delivers what we want,” says Joy, CIO of Control Risks, a $219 million risk consultancy. Nevertheless, he felt that his business needs were best met by turning over management of firewalls and other devices to AT&T, while keeping his antispam function handled in the cloud by a separate e-mail security company. For him, it was a question of one-stop shopping versus what he perceived as best-of-breed.

On a broader scale, it’s unclear whether home consumers will ever want to sign up for a “clean” Internet. AT&T is testing how it could roll out a version of its corporate security offerings to home customers, but already executives have concluded that even its target audience—parents of school-aged children—might not be content with just a Disneyfied version of going online. “Maybe Dad wants to do online gambling but keep teens away from it,” Amoroso says. “We’re just trying to create something people will like and that matches what people want to do.” That will likely involve different versions of the Internet, perhaps delivered to homes based on who’s at the computer—a far cry from really cleaning up the junk in the pipes of the Internet.

For now, and maybe for the long run, companies like AT&T will have to continue to make careful decisions about what traffic they can, safely, delete without violating their service-level agreements with customers or overstepping their bounds as common carriers that just pass bits from left to right. Amoroso says that AT&T can and does delete malicious traffic that will affect its infrastructure. It also deletes e-mail traffic coming from known blacklists of spammers, and blocks port 25 on its DSL lines unless a customer requests otherwise. (Amoroso estimates that 75 percent of spam comes from compromised home PCs, usually on port 25, which is not the port that a typical DSL subscriber uses for outbound e-mail.) But for the most part AT&T can only do so on behalf of a customer—not on behalf of the Internet at large.