Reimagining Security for the Internet of Everything

Over millennia we humans evolved a powerful and personal instinct — trust — that helps to protect us as we make our way through life. It is a vital tool for survival in the physical world and weaves the fabric of our society. When we are in a relationship based on trust we are less vulnerable, which in turn allows us to collaborate and to be creative. Trust is also context specific — you trust your mechanic to fix your car, but probably not to manage your bank account. This is the principle of “need to know”: in each context only information that is needed for that context is available, and nothing more.

What is the meaning of trust in the digital world? The potential for on-line collaboration seems boundless, but here our finely tuned instincts for trust that help to protect us in the physical world are useless: Did that email really come from a colleague? Is the attachment a photo, or a virus? Are you a good enough friend to be my friend on Facebook? Are my Twitter followers admirers or competitors? There is no reliable information to help us decide and so we sometimes make mistakes — clicking on bad links or attachments that give an attacker access to personal or corporate data.

Unfortunately the attacker will probably succeed even if you use the best end-point security software on the market. Why? It is mathematically provable that reliably detecting malware in order to block it — the entire rationale of the security software industry — is impossible. In practice, detection rates for today’s advanced threats are typically around 5-10 percent. We need to accept that, just like us, our computers cannot distinguish good from bad. Anti-virus and other security products that claim to be able to detect malware quite simply cannot keep up.

0 yorum: