Modern virus detection is based on observing patterns either directly by observation or indirectly, usually through some sort of heuristic or intelligent algorithm. These scanning devices are mostly set at identifying a known threat and are often very poor detectives when it comes to evaluating a potential threat. IDS systems have grown in popularity, in part, because they are capable or recognizing patterns which exhibit behavior that resembles some known exploit in one way or another, but they, too, fall short of anticipating the next generation of superbug computer viruses.
The scary fact is that no computer virus yet discovered has successfully capitalized on exponential advances in computer algorithms. Key concepts that are part of everyday programming in the business world have not been demonstrated in the wild by virus writers to date, but they surely will come.
Polymorphism - a key tenet in object oriented programming, will soon be featured by many viruses to adapt to their surroundings. Not only will this allow them to change their appearance or signature, but they will also be able to adapt their behavior according to their environment and erase the pseudo boundaries penned in our imaginations by operating systems, protocols, and other man-made constucts.
Reflection and Interpretation - the ability of a program to inspect its own code, make adjustments and reinvent itself is a building block for intelligent viruses. Using this methodology a malicious programmer could engineer a superbug that can evaluate its behavior against a set of goals and adjust its interworkings accordingly. Such a virus could even incorporate the ability to analyze information feeds, such as public reaction to the virus, efforts to contain or eliminate it, and the operations in the infected system to stay a step ahead of security analysts.
Evolutionary and Genetic Algorithms - As the level of sophistication of viruses goes up, they will be endowed with the ability to mutate and evolve. Genetic algorithms compartmentalize functionality and allow subtle, and often random changes to occur in code. Coupled with an abundance of hosts, these viruses and cancers will mimic the behavior of their biological counterparts; including strains that are active and passive, lethal and symbiotic, and in some cases benign and malignant. Given a large enough playing field (and the internet promises to be just that) these modified strains will compete and indirectly adopt stratagem that ensure their survival. They will cease to be simple programs and come alive.
Intelligence - Perhaps the biggest conundrum is the question of intelligence. It is clear that intelligence will evolve on internet and will increase exponentially over time, even to the point of surpassing our own intelligence.
This is the natural product of any evolutionary system. It will begin with some sense of self awareness, such as reflection, knowledgeable of its own state, and able to make limited adjustments to it. The next generation will be aware of its surroundings, and capable of interpreting that data to adapt. Next will come the awareness of others and the ability to observe and anticipate changes in the environment. Simple rules will then emerge that determine how to operate in a given situation. These rules will lead to swarms of evolving virtual organisms, taking cues from one another, much as bees, ants, fish, and birds do. Communication will follow. Simple signaling will uncover the ability to influence the environment, leadership will emerge, and highly organized groups will work together toward common goals.
The black hats are already at work on these new viruses that will threaten our society, our economy, and our prosperity. The only remaining question is what evolution has in store for humanity. My guess is that organic life will give way to something more stable, either through extinction or adaptation. I just hope that we learn to adapt before our creations realize that they don't need us.
CORE™
0 comments: