SecurityBrief Australia logo
Australia's leading source of cybersecurity and cyber-attack news
Story image

Ransomware attacks on AI algorithms: Risks and solutions

By Sara Barker
Mon 11 Dec 2017
FYI, this story is more than a year old

Artificial intelligence has become a part of our life – its objectively huge potential is now obvious to everyone, more and more is being said about innovative products that give an idea of ​​an AI-operated world of the future, but less so, about the risks associated with the introduction of such technologies – primarily because they are not considered relevant yet.

Nevertheless, while developing solutions for data protection with the introduction of AI, we can confidently say that the relevance of these risks will become a problem not in years, but months from now.

Let's look into it. Starting with the fact that along with promising technological trends - machine learning and AI - the technologies of cyberthreats grow and develop at the same pace, if not faster – recent cases of WannaCry and NonPetya only prove that.

AI algorithms, with all their advantages, have a fundamental problem – data sensitivity. The general weakness with most of the algorithms created so far is that they are trained not to understand information, but to recognize the right answers. Knowing of such algorithm principles allows attackers find ways to deceive them.

There are several potential ways of doing this that we know of. The first one is to apply the poison attack at a training stage. The second is to force algorithm to make the wrong decision using various methods at the stage of application.

We do not present real cases of such attacks, however, desk experiments on images have already been described – when the program mistakes what is in the picture. Same with texts, when the chat bot is forced to give incorrect answers to user’s questions. While this does not pose a serious danger, but as the robots take more and more responsible decisions - and that is inevitable - this can become a big problem.

One of the most interesting cases of such attacks is attacks on computer vision algorithms. A well-known example is that when an image of a panda bear - easily recognizable by a person - is overlaid with "noise" and gradient points form a completely different picture – recognized by an AI as an image of a baboon.

A far more intimidating example was when the same gradient points were used to mask an image of the road “stop” sign - the human eye saw it, but computer vision did not recognize it as a stop sign.

In this case, there is clearly a high probability that an automatic car, having reached the intersection, simply would not stop. Such cases have already been described by investigators, which means that real attacks by the same ransomware is a matter of the nearest future.

On the one hand, being aware of the AI vulnerabilities at an early stage of its development gives scientists the opportunity to eliminate these gaps and make algorithms more effective, but the same knowledge gives cybercriminals the opportunity to improve methods of attacking algorithms and the fact that intruders haven’t used it yet is just a matter of time. The only thing that still deters cybercriminals is access to computing resources.

And what to do with it? In our opinion, the most important step is to process the initial data correctly.

Make sure you use trusted sources

We ourselves use only our own controlled sources and data from our users' computers, which we have the opportunity to verify.

Do a qualitative analysis

Preprocessing of data is needed at all stages, but using it during the training phase will allow to remove anomalies from the original data, reduce the risk of using incorrect data and, consequently, the risk from such attacks. For example, when processing databases with pictures, using them in their original form will make them vulnerable to poisoning attack, while smoothing the image will reduce the probability of such attacks to zero.

One more example. Knowing how the classic antivirus works, malicious attackers regularly re-encrypt the body of a malicious program, changing the hash so that at least for a while it is not recognized by antiviruses that are operating on signatures. We have been familiar with such tricks for a long time – they give us additional information for the AI ​​algorithm, which eventually helps identify the malicious program.

Set the maximum number of situations during the algorithm’s training phase

Every day we train a new model and constantly receive data from Active Protection about what is happening with our customers. We assume that it is possible to uncover our algorithm, come up with a way to attack and start sending us inaccurate data, which would reduce the quality of the algorithm’s performance.

In theory, it could work, but in reality, regularly monitoring the quality of the algorithm and tracking what was yesterday or a week ago, we know how accurate it is. If we see that the accuracy has decreased, we will analyze the sources of new data and understand whether there is a real threat. Constant synchronization and tracking of changes will not prevent the attack itself, but will save you from its consequences.

Do not neglect technical methods

In addition to the correct organization of data processing, do not forget about the technical tools: for example, use the “drop out” method it will allow the network to carry on with its functions correctly even if a certain number of neurons fail. That would provide you with the neural network stability.

The trend of using AI in cybersecurity obliges to pay maximum attention to two important components - high accuracy of ransomware recognition and protecting algorithms from possible attacks. We gave simple examples of how knowing the principles of constructing an algorithm, attackers can find ways to penetrate it.

Therefore, it is important to be aware, to pay maximum attention to the quality of the initial data and to remember that AI can be used not only in cybersecurity, but also in cybercrime - today we can already assume that in the future the malicious programs will be using AI algorithms for their attacks.

Article by Sergey Ulasen, Acronis director of Development, PhD.

Related stories
Top stories
Story image
Ransomware
APAC organisations fail to disclose ransomware breaches
85% of organisations in APAC were breached by ransomware at least once in the past five years, but only 28% publicly disclosed the incident.
Story image
Cybersecurity
Cybersecurity prompts upgrade for 1.3 billion electricity meters
ABI Research finds Advanced Metering Infrastructure (AMI) and cybersecurity concerns are prompting the upgrade of 1.3 billion electricity meters by 2027.
Story image
Training
Trojan cyber attacks hitting SMBs harder than ever - Kaspersky
In 2022 the number of Trojan-PSW detections increased by almost a quarter compared to the same period in 2021 to reach 4,003,323.
Story image
Cybersecurity
Comcast to use ThreatQuotient for cybersecurity operations
Comcast, the parent company of NBC Universal and SKY Group, has chosen ThreatQ Platform and ThreatQ Investigations to meet their cybersecurity needs.
Story image
Ransomware
A third of companies paying ransom don’t recover data - report
Veeam's report finds 76% of businesses who are victims of cyberattacks paid the ransom to recover data, but a third were still unable to get their information back.
Story image
Cybersecurity
BlackBerry offers Kaspersky replacement cybersecurity for the channel
BlackBerry advises that users of Kaspersky software in Australia and New Zealand undertake a rigorous risk analysis of their current security posture.
Story image
Qualys
Qualys updates Cloud Platform solution with rapid remediation
The new update is designed to enable organisations to fix asset misconfigurations, patch OS and third-party applications, and deploy custom software.
Story image
Cybersecurity
Noname Security partners with Netpoleon to target API issues
Specialist API security firm Noname Security has appointed Netpoleon as its distributor in Australia and New Zealand.
Story image
Sift
Sift shares crucial advice for preventing serious ATO breaches
Are you or your business struggling with Account Takeover Fraud (ATO)? One of the latest ebooks from Sift can provide readers with the tools and expertise to help launch them into the new era of account security.
Story image
Application Security
What are the DDoS attack trend predictions for 2022?
Mitigation and recovery are vital to ensuring brand reputation remains solid in the face of a Distributed Denial of Service (DDoS) attack and that business growth and innovation can continue.
Story image
Employment
Tech job moves - Forcepoint, Malwarebytes, SolarWinds & VMware
We round up all job appointments from May 13-20, 2022, in one place to keep you updated with the latest from across the tech industries.
Story image
SaaS
Rubrik Security Cloud marks 'next frontier' in cybersecurity
"The next frontier in cybersecurity pairs the investments in infrastructure security with data security giving companies security from the point of data."
Story image
New Relic
New Relic launches vulnerability management platform
New Relic has introduced New Relic Vulnerability Management to help organisations find and address security risks faster and with greater precision.
Story image
Remote Working
How zero trust and SD-WANs can support productive remote working
The way people connect with applications and data has changed, users are remotely accessing resources that could be stored anywhere from a corporate data center to the cloud.
Story image
Ivanti
Ivanti and Lookout bring zero trust security to hybrid work
Ivanti and Lookout have joined forces to help organisations accelerate cloud adoption and mature their zero trust security posture in the everywhere workplace.
Story image
Cybersecurity
Hard numbers: Why ambiguity in cybersecurity no longer adds up
As cybersecurity costs and risks continue to escalate, CEOs continue to struggle with what their investment in cyber protection buys. Getting rid of ambiguity becomes necessary.
Story image
Artificial Intelligence
How to ensure ethical deployment of AI implementations
The increase in automation and machine technology such as AI and machine learning has unlocked a whole new level of scale and service to organisations. 
Story image
Cybersecurity
More than 40% of banks worried about cloud security - report
Publicis Sapient's new report finds security and the lack of cloud skills and internal understanding of business benefits are big obstacles for banks moving to the cloud.
Story image
Cybersecurity
The 'A-B-C' of effective application security
Software applications have been a key tool for businesses for decades, but the way they are designed and operated has changed during the past few years.
Story image
trust
9/10 Aussies to stop spending if personal data compromised
"Based on the patterns we are seeing among Australian consumers, it is evident that trust in a brand is exceptionally important."
Story image
Cloud Security
Aqua Security createa unified scanner for cloud native security
“By integrating more cloud native scanning targets into Trivy, such as Kubernetes, we are simplifying cloud native security."
Story image
Remote Working
Australia’s remote workers face connectivity and security issues
SOTI's new report finds better video conferencing technology and improved security measures are top concerns for remote workers in Australia.
Story image
Tech job moves
Tech job moves - Datacom, Micro Focus, SnapLogic and VMware
We round up all job appointments from May 6-12, 2022, in one place to keep you updated with the latest from across the tech industries.
Story image
Ransomware
Cybersecurity starts with education
In 2021, 80% of Australian organisations responding to the Sophos State of Ransomware study reported being hit by ransomware. 
Story image
Migration
Let’s clear the cloud visibility haze with app awareness
Increasingly, organisations are heading for the cloud, initiating new born-in-the-cloud architectures and migrating existing applications via ‘lift and shift’ or refactoring.
Story image
Malware
New vulnerabilities found in Nuspire’s Q1 2022 Threat Report
“Threat actors are quickly adjusting their tactics and these exploits tend to get industry attention, but the threat posed by older and attacks still persists."
Story image
VPN
The most common online scams in Australia
No one is safe from online scammers, and many of these scammers have capitalised on the pandemic, using this confusing time to attack more people than ever.
Story image
Nozomi Networks
Nozomi Networks, Siemens reveal software integration
Nozomi Networks and Siemens have extended their partnership by embedding Nozomi Networks’ software into the Siemens Scalance LPE local processing engine.
Story image
Cybersecurity
Infoblox's State of Security Report spotlights Australian remote work hazards
Attackers exploit weak WiFi, remote endpoints, and the cloud, costing 50% of organisations over $1.3 million in breach damages.
Story image
Phishing
KnowBe4 celebrates reaching 50,000 customers worldwide
KnowBe4 has reached the milestone of 50,000 customers, adding nearly 2,500 in the first quarter of 2022 alone.
Story image
Cybersecurity
Managed service providers: effective scoping to avoid costly vendor pitfalls
Managed security services are outsourced services focusing on the security and resilience of business networks.
Story image
Phishing
Google reveals new safety and security measures for users
Google's new measures include automatic two step verification, virtual cards and making it easier to remove contact information on Google Search results.
Story image
BeyondTrust
BeyondTrust integrates Password Safe solution with SailPoint
BeyondTrust has announced the integration of BeyondTrust Password Safe with SailPoint identity security offerings.
Story image
Cybersecurity
Video: 10 Minute IT Jams - An update from IronNet
Michael Ehrlich joins us today to discuss the history of IronNet and the crucial role the company plays in the cyber defence space.
Story image
Vectra AI
Understanding the weight on security leader’s shoulders, and how to shift it
Millions of dollars of government funding and internal budgets are being funnelled into cybersecurity to build resilience against sophisticated threats, indicating how serious this issue has become.
Story image
Silver Peak
The path to an adaptive, modern network
Managing and securing the network looks different than it did just two years ago—especially given that most of these networks are made up of multi-generations of infrastructure stitched together over time.
Story image
Ransomware
Alarming surge in Conti Ransomware Group activity - report
A new report has identified a 7.6 per cent increase in the number of vulnerabilities tied to ransomware in Q1 2022.
Story image
Workato
Workato unveils enhancements to enterprise automation platform
"The extra layer of protection with EKM, zero-logging, and hourly key rotation gives customers a lot more visibility and control over more sensitive data."
Story image
Cybersecurity
A10 Networks finds over 15 million DDoS weapons in 2021
A10 Networks notes that in the 2H 2021 reporting period, its security research team tracked more than 15.4 million Distributed Denial-of-Service (DDoS) weapons.
Story image
Supply chain
Jetstack promotes better security with supply chain toolkit
The web-based resource is designed to help organisations evaluate and plan the crucial steps they need to establish effective software supply chain security.
Story image
Artificial Intelligence
AI-based email security platform Abnormal Security valued at $4B
"A new breed of cybersecurity solutions that leverage AI is required to change the game and stop the rising threat of sophisticated and targeted email attacks."
Story image
Digital Transformation
How to modernise legacy apps without compromising security
At a time when digital transformation has become central to business, even the most important applications come with a ‘use-by’ date.
Story image
Data Protection
Barracuda launches new capabilities for API Protection
"Every business needs this type of critical protection against API vulnerabilities and automated bot attacks," Barracuda says.
Story image
Apricorn
Data backup plans inadequate, data still at risk - study
The Apricorn 2022 Global IT Security Survey revealed that while the majority organisations have data backup plans in place, data for many are at risk.