People today are increasingly reliant on smartphones, smart speakers and other gadgets. Most can’t imagine going more than a few hours without using a computer, and some of them spend most of their work days sitting in front of one. This shift towards a tech-centric culture means people are at a much higher risk of cyberattacks.
Here are five futuristic ways to fight cyberattacks.
1. A high-tech computer chip that proactively prevents attacks
For now, a common way to safeguard against attacks is to make software patches and install them on users’ computers as necessary. Similarly, virus and malware scanners detect suspicious files and keep them quarantined in dedicated folders on a hard drive.
However, researchers at the University of Michigan think they’ve come up with a better way. It’s a computer chip that encrypts and reshuffles its data and coding 20 times per second. As such, even if a hacker breaks into a computer, the information they need to exploit a vulnerability vanishes within milliseconds.
While using a prototype processor fitted with the chip, the people on the research team demonstrated how the tiny component successfully prevented every kind of control-flow hack, which is one of the most commonly used and dangerous attacks hackers carry out.
The rate at which the chip scrambles the data is the “churn,” and it’s possible to adjust its speed. Choosing a churn rate of every 50 milliseconds slowed the processor’s performance by only about 1%, but the churn is several thousand times faster than what any electronic hacking tools accomplish.
2. Compressing network activity to give analysts more of the information they need
Speed is a crucial metric when devising new ways to fight cyberattacks. That’s due, in part, to the significant expense of data breaches.
Research indicates malicious or criminal attacks leading to data breaches are the most costly, resulting in an expense of $157 per user. So, the longer an attack goes undetected, the more expensive the catastrophe becomes.
Researchers working for the US Army believe they found a method that allows detecting harmful network activity sooner than previously used techniques permitted. For example, distributed network intrusion detection tasks a small number of specialty analysts to monitor several networks simultaneously. Sensors on a protected system transmit data to analysis servers, which is a bandwidth-heavy process.
Most systems minimize the bandwidth used by only sending summaries of network traffic. But that means analysts only see snapshots and often spend too much time investigating false positives, or do not have enough details in context to notice genuine attacks.
The researchers hypothesized that malicious network activity manifests early. They developed a tool that stops network transmissions after a predefined number occurs. The next part of the investigation involves compressing traffic analysis to less than 10% of its original volume while sacrificing 1% or less of the cybersecurity alerts.
3. Boosting the cybersecurity of the cloud with blockchain technology
A growing number of businesses are deciding it makes sense to increasingly rely on cloud technology to meet company needs. According to a 2017 poll, 95% of the 1,000 respondents said they were using the cloud. Cloud technology caters to enterprise-level requirements, but it’s not without cybersecurity risks.
Most people know of blockchain technology associated with cryptocurrencies. Information gets verified and permanently added to a digital ledger. As such, it’s difficult to tamper with the content, especially since the blockchain gives visibility and transparency to all involved parties.
Experts insist that making the cloud more secure with the blockchain is not immediately feasible. And, the blockchain is not the sole solution for cloud security, but researchers think it could help propel progress.
4. A human-machine technology to improve cybersecurity accuracy
Many of today’s cybersecurity detection technologies can identify anomalies. When they detect activity that strays from the norm, the systems notify human technicians to take a closer look. A research team from MIT wondered if they might push cybersecurity forward by combining machine learning artificial intelligence (AI) with human intuition. Typically, platforms that use machine learning get smarter over time without input from people.
The technology works by poring over the data and grouping it into clusters through an unsupervised learning process. The goal is for the technology to figure out which strange events are likely cybersecurity attacks. However, the system doesn’t stop there. Next, it provides the clustered data to human analysts. Those people then apply their knowledge and experience when checking the algorithm’s findings.
The humans verify which events are genuine attacks, then give feedback used to make better models for the next set of data. Moreover, the existing models can get better from the updated data in a matter of hours. As such, there is a low to non-existent risk that scientists would rely on outdated algorithms for too long.
5. The first multi-entity detection and response platform
One of the challenging realities of cybersecurity is that risks can come from multiple sources. For example, a person might unknowingly download an attachment contaminated with malware. Or, an adversary could attack the entire network by focusing on a detected flaw. So, one practical cybersecurity approach entails looking for numerous kinds of threats and safeguarding against all of them as much as possible.
Kayla Matthews is a freelance journalist. This article first appeared in the World Economic Forum.