IMSafer scans your kids' IMs for Foleys

A few months ago I wrote about a cute little application called OhDontForget, created by UI designer Jason Stirman. I wondered when he was going to do something serious. He has. Here it is: IMSafer, a new service designed to protect children using instant messaging from online predators.

Part of this solution is software that you download to your computer. It monitors IM traffic for text that's likely coming from (or going to) a predator or pedophile, and sends an email alert to the person who set up the service. IMSafer does not allow parents to spy on everything their child is saying in IM (who has time to read IM logs anyway?), only the context around words and phrases that it flags as potentially dangerous. Parents should also know that IMSafer is also not a totally stealthed application. While the monitor runs very unobtrusively, it can been found and disabled if your child knows how to do such things.

What's clever about the service is that if the software is installed on another computer -- say at a friend's house or at your kid's school -- alerts generated from your child's IM account at that location will be sent to you, not to the owner of the other PC. Likewise, if a friend of your child sets off an alert at your house, their parent will get the alert. Of course, you do have to get other parents and schools to install the software for this to work, and unfortunately there's no Mac version yet. The program also only checks IM traffic from AOL, Microsoft, and Yahoo; support for other services (including IM embedded in social networking sites) should be added soon.

Setting up IMSafer is very easy, and you don't have to know the IM screen names your child has; it tracks all conversations on the PCs you install it on, and then it also watches for those accounts to be used on other PCs that have IMSafer installed. The Web site could offer parents more control over which accounts are managed, however. For example, if your child first logs on to an IM account at another location, you won't get his or her alerts, and it's unclear to me how you could claim this user ID as your child's even if you knew what it was.

I tried the system out by telling a shopping bot I wanted to meet in the real world. It accurately flagged the text and sent me an email alert. Since IMSafer uses lexical analysis, it can't be perfect, but over time it should get even better at spotting dangerous communication. Brandon Watson, IMSafer's founder, is sure that the system would have flagged the explicit IMs that congressman Mark Foley sent to a former page, had it been installed on the page's PC.

The system lets parents vote on alerts that have been sent to other parents, and if enough parents think the alerts are justified, it marks the accounts of the senders of those messages, so that if a person who regularly sends inappropriate messages to other children tries to communicate with yours, you'll get an alert.

IMSafer looks like a good online chaperone, one that will improve as the company tweaks its analysis engine. I don't know of any other system that can keep an eye on kids' IM behavior when they're away from home like this one does.

See also BeNetSafe, which monitors your child's MySpace activity.