Artificial intelligence promises to bring a new era to cybersecurity. It will allow the bad guys to have ever-better tools to perpetrate attacks, and it will allow the good guys to have ever-better tools to counteract them. This is an old struggle that dates as far back as cryptography. There is however one area where this may not happen.
Everyday, ten million new devices are plugged into the Internet; devices that aren’t computers, and for the most part are not upgradeable. What this means is that if a security vulnerability could be detected, which may be easier in the future with better AI’s that look for them, it could open a whole host—millions and millions of devices—to security risks that cannot be simply patched or repaired. This could be cameras or ovens or any device that people have attached to the Internet. There’s no easy solution to this, and the fact that Internet devices develop in a kind of ad hoc fashion, there isn’t any overarching strategy to deal with it in the future.