Apr 13, 2021
This is a book that an ICS security professional should give to friends and family to read so they know why they do what they do. Nicole guides the lay person through her compelling journey to understand the 0day market and its impact on the security of the systems we all rely on. The ICSsec pro will find it to be interesting except for the parts on ICS / critical infrastructure where it is a historical fiction ... historical incidents extrapolated to their most dire possible results rather than presented in their true context.
This was a very difficult book to review. I'm conflicted because the story is engaging and will keep the lay person turning the pages. The 0day market through line is well told, and the theme and major points Nicole is making are clear and compelling. And yet the parts I know in detail, ICS security and critical infrastructure, are portrayed in a light that is misleading, and even deliberately misleading. Misleading because a lay reader, including government policymakers, would almost certainly conclude the US critical infrastructure at this moment is compromised and a click or two away from the Russians, or other adversaries, causing a major catastrophe.
The Story ... The Positive
Imagine you are trying to tell your Mom or Dad, your Husband/Wife/Partner, your close friend about cybersecurity risk and how it could affect their lives, their communities and their country or region. This is hard. If you get into the technical details you will lose them. If you try to add too much nuance (HT: RLee) you will lose them. There needs to be a captivating story that makes the non-technical audience keep reading even if they don't care about the tech.
Nicole has succeeded in this area by inserting herself into the story. She is not the heroine of the story. Instead she is the observer, the Nick Carraway from The Great Gatsby, who is observing the players in the 0day world who are neither pure heroes nor pure villains. She begins her journey naively at S4x13 in Miami Beach and investigates for over seven years. Never actually reaching a point of knowing the market, and yet she describes what she knows and what cannot be known.
The best parts of the book are when Nicole is an active character, walking through the world and talking to the players. You feel her frustration, fear, intimidation, dread and disgust. You want her to come out the other side with some answers or even the answer. Although to her credit she does not force a solution. The ending is actually more muddled than the beginning. It makes for a less satisfying journey, and it is more accurate.
Nicole's journey is the highlight of the book. The reason why you can recommend it to your Mom or Dad. It would have been even better for the lay reader had she not tried to add in the history, the details. It does not go deep into the technical details like Kim Zetter's Countdown to Zero Day, which can be viewed as a positive or negative. My view is if you are not going to push for technical accuracy, then less is better. Still the book is an interesting read as my family members can attest.
I'm sympathetic with the theme that the US Government's focus on offense, in this book primarily the accumulation of 0days, has made the world more dangerous. We see this offense focus clearly and unapologeticly stated by NSA and Cyber Command across multiple leaders. The dominance of offensive theory and capabilities makes for a less stable world.
My hope for any policy makers reading this book is they reject the current philosophy of "we can't defend so we need to be able to attack first and potentially cause even greater damage". Nicole repeatedly shows where US actions to buy 0days resulted in an unexpected and negative result.
What is less certain is whether the 0day market was inevitable whether the US participated, or even led, in the early years. Nicole bemoans that "the cyberarms market was an incoherent mess". There were buyers and sellers reaching agreement, so it was not an incoherent mess. It was unregulated and led to undesirable outcomes in the past and likely in the future. However unless there are agreed upon cyber norms, similar to biological and chemical weapons, this was and is to be expected.
The Technical ... The Negative
I'm only qualified to comment on the ICS / Critical Infrastructure part of the book. My guess though is if you are part of the Vulnerability Equities Process (VEP), 0day market, Ecko Party, ... the parts of the book that discuss your area will be frustrating. I say this because anyone reading the ICS / Critical Infrastructure part of the book would come out with an incorrect understanding of the current capability of Russia and other adversaries to cause a catastrophic event using existing deployed exploits of the US critical infrastructure.
There is not a lot of factual detail in the book, again good for the lay person reader, and therefore creating an errata list wouldn't be a compelling case. In the ICS area, there was one major mistake on page 297:
It was an act of unprecedented digital cruelty, but the Russians stopped just short of taking lives. Six hours later, they flipped the power back on in Ukraine, just long enough to send their neighbor, and Kyiv's backers in Washington a clear message: "We can torch you".
This clearly implies that the Russians stopped their attack and turned the power back on in Ukraine. What actually happened was the Ukrainians went out to the substations and manually brought them back on line and operated them manually for many months. The SCADA system was down for about a year. Nicole was right that a "clear message" was sent.
This error on its own in a 400-page book would not be an issue. The issue is that every incident is presented in its worst possible light. Often not wrong by a strict parsing of the text, but misleading. A great example is Wolf Creek Nuclear plant on page 397:
the Russians were inside our nuclear plants ... The code made clear that Russia's hackers had breached the most alarming target of all: Wolf Creek, the 1200 mega-watt nuclear power plant near Burlington, Kansas. This was no espionage attack. The Russians were mapping out the plant's networks for a future attack; they had already compromised the industrial engineers who maintain direct access to the reactor controls ... And the goal wasn't to stop the boom. It was to trigger one.
Although she doesn't state it, this quote and the surrounding text would almost certainly be read as the Russians were in the nuclear control and safety systems. The reality is that an adversary had breached the office network at the Wolf Creek Nuclear Power Plant, but they had not yet been able to breach the ICS that controlled the nuclear plant nor the safety systems that would need to fail to cause "the boom".
Nicole wrote on page 392, "The technical community will argue I have overgeneralized and oversimplified, and indeed, some of the issues and solutions are highly technical and better left to them." When I had my interview with Nicole and wrote this review, this sentence kept running through my mind. After much introspection and consideration of this point, I do believe that this Wolf Creek example and many others in the book would lead the lay person to an incorrect understanding of the current state.
How different would a reader's understanding be if the Wolf Creek incident would have said the Russians were just outside the control and safety systems. Yes, they were knocking on the doors where accounting, HR, and other office functions take place, but they had not yet gotten in to plant operations or safety systems.
Another specific example is related to the Bowman Avenue Sluice Gate. To her credit Nicole notes in an early section that this is not Arthur R. Bowman dam in Oregon. However in the concluding chapter she writes,
"We've caught Iranian hackers rifling through our dams."
An Internet connected, ~5 meter wide, ~1 meter high sluice gate that keeps a neighborhood from flooding a couple of times a year is not a national security event and not worth noting as a reason for perilous concern in the concluding chapter.
Beyond the ICS security specifics, and probably more important, are the unsubstantiated contentions that the Russians and adversaries are in our systems and a click away from causing a catastrophic event. There are many in the book's text and in the interviews.
There are many more examples where the book's clear message is that the adversaries, Russians, Chinese, North Koreans, Iranians are able to cause a critical infrastructure catastrophe. The facts don't indicate this. As noted in the summary, Nicole has taken historical incidents and either extrapolated them to their most hysterical or left out the a sentence or two that would give the reader the correct impression. This approach is consistent throughout the text.
If the goal is to grab the lay reader by the shoulders and shake them saying this is important, it is a successful deception. Still it is nearly as scary without the hyperbole.
The final chapter includes a set of recommendations that are underwhelming. Vendors need to have a security development lifecycle (SDL) and put out better systems. The end users, the people need to be more security aware. In this area I don't fault Nicole because there are not easy answers. It might have been better to leave this chapter off.
One interesting suggestion was on Page 398:
We could start by passing laws with real teeth that mandate, for instance, that critical infrastructure operators refrain from using old, unsupported software; that they conduct regular penetration tests, that they don't reuse manufacturers' passwords; that they turn on multifactor authentication; and that they airgap the most critical systems.
This is NERC CIP, sans the air gap, that has been around for a decade plus.
If you've made it to the end of this book review, I hope you understand where the book succeeds and fails. Who it is written for, and who it is not written for. You and I are not the intended audience. The journey is compelling; the themes are on target; and maybe we should not get too upset that the specifics go beyond reality and are taken to their most extreme possibility.