According to Dan Rayburn, AT&T (NYSE:T, news, filings) has turned away from its in-house CDN development and worked out a deal with EdgeCast Networks (news) [a subsidiary of Verizon (NYSE:VZ, news, filings)] instead. They'll use EdgeCast's software to build out their retail offering. If AT&T runs with it, this could be a very significant alliance for EdgeCast, potentially bringing it fully into the top tier of the CDN market. The other high profile carrier partners they've added are Global Crossing, Deutsche Telekom, Telus, and AAPT. But for sheer size and scope, there is only one AT&T. Yet we really know little about just how big AT&T's CDN business might be or how fast it is growing. They've been at it for years now, but quietly.
This is the second time lately where AT&T chose to outsource content-related technology. Back in October their growing relationship with Cotendo for value added services was revealed, which I still see as less of an attack on Akamai and more defense of its favored position amongst large enterprises. This news about EdgeCast seems to further that hypothesis. AT&T isn't invading the CDN space, rather it's trying to hold onto its customers as the product mix those customers need shifts toward content. See, the content space is invading *their* turf, and they are adjusting in response.
It also suggests that perhaps AT&T won't be following Verizon into the carrier neutral data center and cloud space, or at least not in the same form. They seem less interested in vertically integrating at the moment, and more interested in piecing together stuff that works. EdgeCast still needs to be careful though, being the junior partner to AT&T has risks of its own.