Would peering have helped Sony?
Sony suffered a significant penetration, allowing unauthorized access to privileged information. According to the New York Times, the hackers stole the systems administrator credentials and passwords. Then the contracts, salaries, emails, budgets, medical records, social security numbers, etc. were published and 75% of Sony servers and data were destroyed. No amount of peering would have protected against inherent security vulnerabilities.
What peering can do is architecturally segregate traffic, a common security technique, and peering also establishes a direct escalation path with peered networks to deal with technical issues related to traffic exchange.
I use the word “architecturally” because once established, nothing is actively required to maintain the traffic optimization. If peering goes away, there is always transit to back up to. By the way, large financial institutes like Visa peer because it architecturally improves security.
Traffic segregation is also helpful during times of Denial of Service (DoS) and Distributed Denial of Services (DDoS) attacks. If you contact an ISP Network Operations Center and you aren’t a customer or a peer, they most likely won’t take your call. So we have seen that peering is helpful when coordination is helpful. Some peering coordinators view this coordination feature as the strongest reason to peer.
Traffic segregation can also be helpful for traffic management. Peering allows you to direct and control traffic the way that transit can not. The largest content companies in the world peer to manage and improve the end-user experience. Peering also scales very well - one can aggregate traffic across public peering fabrics and the peaks of one flow can offset the valleys in other flows.