Here are the predictions that came true—and one that didn't.
Like almost every other columnist and pundit in the world, I like to end the year with a set of predictions about what we can expect in the coming year. And since my area of expertise is technology and government, that’s what I concentrate my predictions on. I base them on everything new I learned about cybersecurity over the previous year, as well as conversations with many experts in the field. I’ll be doing this again this year, with my 2018 predictions due around Christmas for those of you won’t yet be in full holiday mode.
Prediction columns are relatively easy for most people because there is little in the way of auditing. But unlike about 90 percent of other columnists, I go back every year and grade how my predictions stacked up. Full disclosure, based on my previous predictions, my pundit accuracy score is 75 percent. That’s not bad for predicting the future of technology and government, though I hope to improve on that accuracy.
So, let’s look at my 2017 predictions and find out how I did.
Prediction 1: Congress will act to secure the internet of things
Oh man, what did I expect predicting that the current Congress would actually do something useful for the world? Last year, we got to see the dangers of having millions of unsecured devices—everything from video cameras to thermostats to smart televisions—accessing the internet with no oversight and no protections. Smart hackers gathered information about millions of those so-called internet of things (IoT) devices and then linked them into a humongous botnet call Mirai. Using Mirai, the hackers unleashed one of the largest denial-of-service attacks ever generated, shutting down some of the most popular sites on the internet for several hours.
After the initial shock and awe of that attack wore off, the hackers doubled down on their efforts, releasing the source code for Mirai, which would presumably allow anyone with a little bit of programming skill to string their own botnets together. Amid that crisis, calls went out to legislators asking for new laws requiring that security be a part of all future IoT devices.
There has been some movement in this area. In August, Sen. Mark R. Warner, D-Va., introduced the Internet of Things Cybersecurity Improvement Act of 2017, which requires the government to secure its IoT devices. I was hoping for something more like HIPAA requirements, where manufacturers would be held responsible if they deployed unsecured devices which were later used in an attack. Still, getting the government to fix its networks first is a good start.
Now if we could just get Congress to vote on the bill, we might get somewhere. As of the last update, S.1691 has been read twice and referred to the Committee on Homeland Security and Governmental Affairs where it sits and sits and sits.
Industry does seem to be stepping up, however. The Cloud Security Alliance has updated their guidelines on how to develop and secure future IoT devices. And most cybersecurity programs I reviewed in 2017 included at least some ability to secure IoT. Also, there have been no new major IoT attacks, so perhaps protection is slowly catching up.
Verdict: Ongoing. Yes, that is a bit of stretch, and I should say incorrect, but technically, Congress still has a little time to act before the end of the year to act— unless they are too busy concentrating on other things.
Prediction: Despite the danger, there will be no major election reform.
I was fairly confident that nobody would be able to hack a presidential election and said as much in my pre-election column last year. The reason is not because of robust cybersecurity, but instead because every state and local government handles their own voting system. Many of the systems are hopelessly outdated, with quite a few existing in closed network environments, meaning it would take an army of hackers working together to pull off a major voting attack. Accomplishing that and not getting caught would be even more impossible.
Evidence has been uncovered regarding foreign involvement in the election, but it is all peripheral to the actual voting systems. Agents working for Russia trolled Facebook and bought ads that reportedly reached 126 million people. Whether those ads illegally helped Donald Trump win is an open question, but for our purposes, the fact that Russia had to resort to trolling on Facebook probably means that they had no real footprint into the actual voting systems.
And it's not like we aren’t taking election security seriously. The National Institute of Standards and Technology, working with the U.S. Election Assistance Commission, developed technical Voluntary Voting System Guidelines to help states prepare good cyber defenses. And the Homeland Security Department offered to monitor state election security, help that was accepted by 30 states for the last election.
Even so, with peripheral Russian meddling already uncovered, many are wondering if we need to do more to secure our right to vote. My prediction, however, was that nothing would be done about it in 2017. And sadly, this was correct. The beneficiary of the election manipulation, in whatever form it took, won the day. So, there is little incentive to go back and investigate the process. It might be a long time before we finally get serious about protecting our elections.
Verdict: Correct! Though I wish it wasn’t so.
Prediction: Endpoint security takes the spotlight.
Last year, I started to see a shift in the types of cybersecurity attacks levied at both the private sector and government. Instead of going directly for core assets, hackers began to more actively target endpoints. In a way, that makes sense. Endpoint security has historically been consigned to just signature-based anti-virus, a technology that has proved inadequate against advanced threats. Once you control an endpoint, you can gather intelligence about core defenses, or even use it as a jumping off point to bigger things.
Thankfully, security industry professionals saw the same trends as I did, and responded with robust new endpoint security, either as modules to existing core security programs or as endpoint-exclusive suites. Many of the programs I reviewed in 2017 were designed to protect those increasingly targeted endpoints.
Those programs include the Promisec PEM (Promisec Endpoint Management) software, which is rare in that it can be installed completely on-premises compared with most of these types of programs that are deployed as services. Security-conscious federal agencies may be more accepting of an on-premises solution. And then there is Minerva, which is designed to shore-up the flaws of traditional anti-virus on endpoints with innovative technologies like deception. The idea is that most normal threats will be blocked by traditional antivirus, and Minerva will detect and kill anything that attempts to get around that protection.
Verdict: Correct! And this trend will continue into 2018.
John Breeden II is an award-winning journalist and reviewer with over 20 years of experience covering technology. He is the CEO of the Tech Writers Bureau, a group that creates technological thought leadership content for organizations of all sizes. Twitter: @LabGuys