The era of AI hacking is closer than you think

1 year ago

And finally, sophistication: AI-assisted hacks open the door to complex strategies beyond those that can be devised by the human mind unaided. Sophisticated statistical analysis of AI can reveal relationships between variables and thus possible exploits that the best strategists and experts would never recognize. This sophistication could allow the AI ​​to deploy strategies that undermine multiple layers of the target system. For example, an AI designed to maximize the share of a political party’s vote can determine the exact combination of economic variables, campaign messages, and voting procedural settings that can make the difference between electoral victory and defeat, expanding the revolution that mapping software has brought in shenanigans. all aspects of democracy. And that’s not even to mention the hard-to-find tricks that AI can offer to manipulate the stock market, legal systems, or public opinion.

With the speed, scale, scope, and complexity of computers, hacking will become a problem that we, as a society, can no longer handle.

I remember a scene from a movie Terminator, in which Kyle Reese describes to Sarah Connor the cyborg that is after her: “You can’t bargain with him. This is impossible to argue. He feels no pity, no remorse, no fear. And it will absolutely never stop…” We are not dealing with literal killer cyborgs, but as AI becomes our adversary in the world of social hacking, it may be just as difficult for us to keep up with its inhuman ability to hunt down our vulnerabilities.

Some AI researchers worry about the extent to which powerful AIs will be able to overcome the limitations imposed by humans and – potentially – become dominant in society. While this may seem like wild speculation, it is a scenario worth at least considering and preventing.

However, today and in the near future, the hacker attacks described in this book will be carried out by the powerful of this world against all of us. All AI, whether it’s on your laptop, on the internet, or embodied in a robot, is programmed by other people, usually for their benefit, not yours. While an Internet-connected device such as Alexa may imitate the role of your trusted friend, never forget that it is meant to sell Amazon products. And just as the Amazon website encourages you to buy your own brands over the competition’s higher quality products, it won’t always work in your best interest. This will break your trust in Amazon for the sake of its shareholders’ goals.

In the absence of any meaningful regulation, there is really nothing we can do to prevent AI hacking from spreading. We must recognize that this is inevitable and build robust governance structures that can respond quickly and effectively to normalize beneficial system breaches and neutralize malicious or unintentionally damaging ones.

This issue raises deeper and more complex questions than how AI will evolve or how institutions might respond to it: what kind of hacks are considered useful? Which ones cause damage? And who decides? If you think a government should be small enough to drown in a bathtub, then you’re probably thinking that hacks that reduce the government’s ability to control its citizens are usually good. But you still may not want to replace political overlords with technological overlords. If you believe in the precautionary principle, you need as many experts as possible to test and evaluate hacks before they are included in our social systems. And you might want to apply this principle upstream to the institutions and structures that make these hacks possible.

Questions continue. Should AI-generated hacks be regulated locally or globally? Administrators or referendum? Or can we let the market or civil society groups decide? (Current attempts to apply governance models to algorithms are an early indicator of how this will play out.) The governance structures we are developing will give some people and organizations the power to define the hacks that will define the future. We need to make sure that this power is used wisely.

Extracts from Mind of a hacker: how the powers that be change the rules of society and how to change them Bruce Schneier. Copyright © 2023 Bruce Schneier. Used with permission from the publisher, WW Norton & Company, Inc. All rights reserved.

Leave a Reply