As the Obama era comes to a close, Dark Reading asks industry leaders to weigh in on the best and worst of the administration’s cybersecurity policies.
In countering the cyber threat, or any national security challenge, the President’s powers can be divided into two buckets. The first contains actions and policies intended to change the behavior of our adversaries. The second contains actions and policies intended to change the behavior of those we seek to protect. In other words, the President can try to stop the bad guys from attacking us or encourage good guys to take steps to reduce their exposure. Unfortunately for cybersecurity professionals tasked with defending sensitive data and systems, the Obama administration has come up short in both areas.
Ineffective Use of US Power to Create a Deterrent
In the first category, the Obama administration has been too cautious in taking steps to create a real deterrent to nation states who use cyber tools and weapons against us. Take China, for example. The US intelligence community has been warning the President for years that China had embarked on a strategic cyber espionage campaign targeting US companies. Despite these warnings, the administration did little or nothing to dissuade the Chinese from deploying large numbers of hackers to steal intellectual property and business information. In the eyes of many experts, this effort cost US companies’ contracts, and - ultimately - jobs to Chinese state-owned competitors.
The administration appeared to take a bold step in May 2014 by green-lighting the Department of Justice’s indictment of five officers of the Chinese People’s Liberation Army on cyber espionage charges. But this was after many years of sitting on the evidence, hoping softer measures would convince China to change its ways. While the indictments got China's attention, it wasn't until President Obama called out China's President Xi Jinping in September 2015 that we saw any real response from the Chinese. President Xi's pledge not to conduct cyber espionage for commercial gain seemed to signal progress, but the damage had already been done. After years of successfully stealing US intellectual property and business information, China had what it needed to elbow US competitors out of major markets.
A similar stance was taken with the Iranians. Despite Iran’s aggressive use of distributed denial of service (DDoS) attacks against the US financial sector in 2012 and 2013, the administration did little or nothing to make Iran pay a real price for its aggression. It was not until after the nuclear framework with Iran was concluded in July 2015 that the administration finally called them out on their cyberattacks. This came in the form of an indictment of seven Iranian hackers for the attacks on US banks’ websites and a dam in New York. The timing of the indictment again diluted its deterrent effect.
The list of missed opportunities to create a deterrent doesn’t end there. While there may be much going on behind the scenes, it seems the administration has let a number of significant cyberattacks go unanswered. For example, other than weak and ineffective trade sanctions, North Korea appears to have paid no price for the devastating attack on Sony in 2014. Similarly, the more recent attacks against the DNC and other networks related to the presidential campaign, purportedly by the Russians, have been absorbed without a significant response.
Nation-state actors are (mostly) rational. They conduct a risk-reward analysis before taking on the US with cyberattacks or any other aggressive behavior. The administration has not made our cyber adversaries pay a price, so cyberattacks are little risk and great reward. We have to change the inputs on the risk side to change their behavior.
Ineffective Policies to Encourage Self-Defense
Compounding the administration’s hesitant use of US power to create a deterrent is its myopic focus on “information sharing” as the bedrock of cyber protective strategies. While important, the strict focus on information sharing as a countermeasure is misplaced for two reasons. First, a lack of information sharing doesn’t seem to be the problem. We see from studies like the Verizon Data Breach Investigations Report (2015) that 99.9% of the vulnerabilities exploited by hackers were known for more than a year. This is a staggering statistic. It shows the problem for enterprises is not sharing information, it’s effectively using the information (e.g. patching the vulnerabilities) that we already have.
The value of “information sharing” as the foundation of a cyber defense strategy depends almost entirely on the type of information shared. If the information takes the form of a “signature” of “Indicator of Compromise,” (IOC), a defender may be able to configure defenses to detect such activity. But this approach has been demonstrated to have limited value, especially against sophisticated adversaries.
The administration is well aware of the limitations of signature-based systems because it owns and operates one. Specifically, the National Cybersecurity Protection System, operationally known as “Einstein,” is an integrated system-of-systems that is intended to deliver a range of capabilities, including intrusion detection, intrusion prevention, analytics, and information sharing. In January of 2016, the US Government Accountability Office (GAO) issued a report that criticized Einstein with facts that cybersecurity professionals know all too well.
According to the GAO report, Einstein “provides DHS with a limited ability to detect potentially malicious activity entering and exiting computer networks at federal agencies. Specifically, [Einstein] compares network traffic to known patterns of malicious data, or “signatures,” but does not detect deviations from predefined baselines of normal network behavior. In addition, [Einstein] does not monitor several types of network traffic and its “signatures” do not address threats that exploit many common security vulnerabilities and thus may be less effective…”
Disincentives for Sharing: What Went Wrong
Information sharing as a basis for national strategy falls short for a second reason – we are not sharing all of the information available. This is because there are significant disincentives for organizations to open up about what they did wrong or how they let a breach happen. This type of information is potentially far more useful for network defenders.
For example, it would help to know what Target actually did or failed to do in the days leading up to their now famous breach in November 2013. But we may never know the facts - even though they could help others to avoid repeating the same mistakes - because sharing such information is an invitation to litigation. The Target breach does not stand alone in this regard. In the current legal framework, some of the most valuable information for cyber defense never sees the light of day because there are real disincentives for sharing.
Despite the limited value of sharing signatures and IOCs, the administration has not been successful in creating incentives and removing penalties to expand the types of information being shared. This is arguably the result of resistance from the private sector and Congress. The net effect, however, is a continued focus on technical indicators that make signature-based systems, like Einstein, less effective.
I believe the administration tried hard, and would have changed the rules to encourage more effective sharing if they could, but they just couldn’t get it done. I also believe that nation states don’t care how hard we try. Their only concern is whether they can achieve their goals with within their cost and risk tolerance parameters. After seven years of the administration’s best efforts, the fact remains that nation state actors are seeing more upside than ever.