Cyber security is a manpower constrained market – therefore the opportunities for AI automation are vast. Frequently, AI is used to make certain defensive aspects of cyber security more wide reaching and effective: combating spam and detecting malware are prime examples. On the opposite side there are many incentives to use AI when attempting to attack vulnerable systems belonging to others. These incentives could include the speed of attack, low costs and difficulties attracting skilled staff in an already constrained environment.
Current research in the public domain is limited to white hat hackers employing machine learning to identify vulnerabilities and suggest fixes. At the speed AI is developing, however, it won’t be long before we see attackers using these capabilities on mass scale, if they don’t already.
How do we know for sure? The fact is, it is quite hard to attribute a botnet or a phishing campaign to AI rather than a human. Industry practitioners, however, believe that we will see an AI-powered cyber-attack within a year: 62% of surveyed Black Hat conference participants seem to be convinced in such a possibility.
Many believe that AI is already being deployed for malicious purposes by highly motivated and sophisticated attackers. It’s not at all surprising given the fact that AI systems make an adversary’s job much easier. Why? Resource efficiency point aside, they introduce psychological distance between an attacker and their victim. Indeed, many offensive techniques traditionally involved engaging with others and being present, which in turn limited attacker’s anonymity. AI increases the anonymity and distance. Autonomous weapons is the case in point; attackers are no longer required to pull the trigger and observe the impact of their actions.
It doesn’t have to be about human life either. Let’s explore some of the less severe applications of AI for malicious purposes: cybercrime.
Social engineering remains one of the most common attack vectors. How often is malware introduced in systems when someone just clicks on an innocent-looking link?
The fact is, in order to entice the victim to click on that link, quite a bit of effort is required. Historically it’s been labour-intensive to craft a believable phishing email. Days and sometimes weeks of research and the right opportunity were required to successfully carry out such an attack. Things are changing with the advent of AI in cyber.
Analysing large data sets helps attackers prioritise their victims based on online behaviour and estimated wealth. Predictive models can go further and determine the willingness to pay the ransom based on historical data and even adjust the size of pay-out to maximise the chances and therefore revenue for cyber criminals.
Imagine all the data available in the public domain as well as previously leaked secrets through various data breaches are now combined for the ultimate victim profiling in a matter of seconds with no human effort.
When the victim is selected, AI can be used to create and tailor emails and sites that would be most likely clicked on based on crunched data. Trust is built by engaging people in longer dialogues over extensive periods of time on social media which require no human effort – chatbots are now capable of maintaining such interaction and even impersonate the real contacts by mimicking their writing style.
Machine learning used for victim identification and reconnaissance greatly reduces attacker’s resource investments. Indeed, there is even no need to speak the same language anymore! This inevitably leads to an increase in scale and frequency of highly targeted spear phishing attacks.
Sophistication of such attacks can also go up. Exceeding human capabilities of deception, AI can mimic voice thanks to rapid development in speech synthesis. These systems can create realistic voice recordings based on existing data and elevate social engineering to the next level through impersonation. This, combined with other techniques discussed above, paints a rather grim picture.
So what do we do?
Let’s outline some potential defence strategies that we should be thinking about already.
Firstly and rather obviously, increasing the use of AI for cyber defence is not such a bad option. A combination of supervised and unsupervised learning approaches is already being employed to predict new threats and malware based on existing patterns.
Behaviour analytics is another avenue to explore. Machine learning techniques can be used to monitor system and human activity to detect potential malicious deviations.
Importantly though, when using AI for defence, we should assume that attackers anticipate it. We must also keep track of AI development and its application in cyber to be able to credibly predict malicious applications.
In order to achieve this, a collaboration between industry practitioners, academic researchers and policymakers is essential. Legislators must account for potential use of AI and refresh some of the definitions of ‘hacking’. Researchers should carefully consider malicious application of their work. Patching and vulnerability management programs should be given due attention in the corporate world.
Finally, awareness should be raised among users on preventing social engineering attacks, discouraging password re-use and advocating for two-factor-authentication where possible.
The Malicious Use of Artificial Intelligence: Forecasting, Prevention, and Mitigation 2018
Cummings, M. L. 2004. “Creating Moral Buffers in Weapon Control Interface Design.” IEEE Technology and Society Magazine (Fall 2004), 29–30.
Seymour, J. and Tully, P. 2016. “Weaponizing data science for social engineering: Automated E2E spear phishing on Twitter,” Black Hat conference
Allen, G. and Chan, T. 2017. “Artificial Intelligence and National Security,” Harvard Kennedy School Belfer Center for Science and International Affairs,
Yampolskiy, R. 2017. “AI Is the Future of Cybersecurity, for Better and for Worse,” Harvard Business Review, May 8, 2017.
Let’s talk about applying the SABSA framework to design an architecture that would solve a specific business problem. In this blog post I’ll be using a fictitious example of a public sector entity aiming to roll-out an accommodation booking service for tourists visiting the country.
To ensure that security meets the needs of the business we’re going to go through the layers of the SABSA architecture from top to bottom.
Start by reading your company’s business strategy, goals and values, have a look at the annual report. Getting the business level attributes from these documents should be straightforward. There’s no need to invent anything new – business stakeholders have already defined what’s important to them.
Every single word in these documents has been reviewed and changed potentially hundreds of times. Therefore, there’s usually a good level of buy-in on the vision. Simply use the same language for your business level attributes.
After analysing the strategy of my fictitious public sector client I’m going to settle for the following attributes: Stable, Respected, Trusted, Reputable, Sustainable, Competitive. Detailed definitions for these attributes are agreed with the business stakeholders.
Next step is to link these to the broader objectives for technology. Your CIO or CTO might be able to assist with these. In my example, the Technology department has already done the hard job of translating high-level business requirements into a set of IT objectives. Your task is just distill these into attributes:
Now it’s up to you to define security attributes based on the Technology and Infrastructure attributes above. The examples might be attributes like Available, Confidential, Access-Controlled and so on.
The next step would be to highlight or define relationships between attributes on each level:
These attributes show how security supports the business and allows for two-way tracebility of requirements. It can be used for risk management, assurance and architecture projects.
Back to our case study. Let’s consider a specific example of developing a hotel booking application for a public sector client we’ve started out with. To simplify the scenario, we will limit the application functionality requirements to the following list:
|P001||Register Accommodation||Enable the registration of temporary accommodations available|
|P002||Update Availability||Enable accommodation managers to update availability status|
|P003||Search Availability||Allow international travellers to search and identify available accommodation|
|P004||Book Accommodation||Allow international travellers to book accommodation|
|P005||Link to other departments||Allow international travellers to link to other departments and agencies such as the immigration or security services (re-direct)|
And here is how the process map would look like:
There are a number of stakeholders involved within the government serving international travellers’ requests. Tourists can access Immigration Services to get information on visa requirements and Security Services for safety advice. The application itself is owned by the Ministry of Tourism which acts as the “face” of this interaction and provides access to Tourist Board approved options. External accommodation (e.g. hotel chains) register and update their offers on the government’s website.
The infrastructure is outsourced to an external cloud service provider and there are mobile applications available, but these details are irrelevant for the current abstraction level.
From the Trust Modelling perspective, the relationship will look like this:
Subdomain policy is derived from, and compliant with, super domain but has specialised local interpretation authorised by super domain authority. The government bodies act as Policy Authorities (PA) owning the overall risk of the interaction.
At this stage we might want to re-visit some of the attributes we defined previously to potentially narrow them down to only the ones applicable to the process flows in scope. We will focus on making sure the transactions are trusted:
Let’s overlay applicable attributes over process flows to understand requirements for security:
Now it’s time to go down a level and step into more detailed Designer’s View. Remember requirement “P004 – Book Accommodation” I’ve mentioned above? Below is the information flow for this transaction. In most cases, someone else would’ve drawn these for you.
With security attributes applied (the direction of orange arrows define the expectation of a particular attribute being met):
These are the exact attributes we identified as relevant for this transaction on the business process map above. It’s ok if you uncover additional security attributes at this stage. If that’s the case, feel free to add them retrospectively to your business process map at the Conceptual Architecture level.
After the exercise above is completed for each interaction, it’s time to go down to the Physical Architecture level and define specific security services for each attribute for every transaction:
At the Component Architecture level, it’s important to define solution-specific mechanisms, components and activities for each security service above. Here is a simplified example for confidentiality and integrity protection for data at rest and in-transit:
|Service||Physical mechanism||Component brands, tools, products or technical standards||Service Management activities required to manage the solution through-life|
|Message confidentiality protection||Message encryption||IPSec VPN||Key management, Configuration Management, Change management|
|Stored data confidentiality protection||Data encryption||AES 256 Disk Encryption||Key management, Configuration Management, Change management|
|Message integrity protection||Checksum||SHA 256 Hash||Key management, Configuration Management, Change management|
|Stored data integrity protection||Checksum||SHA 256 Hash||Key management, Configuration Management, Change management|
As you can see, every specific security mechanism and component is now directly and traceable linked to business requirements. And that’s one of the ways you demonstrate the value of security using the SABSA framework.
I’ve spend last week in Vienna at the annual intergovernmental conference focused on protecting critical energy infrastructure.
The first two days were dedicated to the issues of security and diplomacy.
A number of panel discussions, talks and workshops covered the following topics:
- Implementing the EU strategy for safe, open and secure cyberspace
- Cyber-threats to critical energy infrastructure
- Operational resilience
- Reducing the risks of conflicts stemming from the use of cyber-capabilities
- Cyber-diplomacy: developing capacity and trust between states
For the rest of the conference we moved from the Diplomatic Academy of Vienna to Tech Gate, a science and technology park and home to a number of local cyber startups.
We’ve discussed trends in technology and cyber security, participated in Cyber Range simulation tutorial and a scenario-based exercise on policy development to address the growing cyber-threat to the energy sector.
AIT Austrian Institute of Technology together with WKO Austrian Economic Chambers, ASW Austrian Defence and Security Industry, and the Austrian Cyber Security Cluster hosted a technology exhibition of latest solutions and products as well as R&D projects.
Participants had an opportunity to see state-of-the-art of next generation solutions and meet key experts in the field of cyber security for protecting critical infrastructures to fight against cyber-crime and terrorism.
Talks continued throughout the week with topics covering:
- Securing the energy economy: oil, gas, electricity and nuclear
- Emerging and future threats to digitalised energy systems
- Cyber security standards in critical energy infrastructure
- Public sector, industry and research cooperation in cyber security
- Securing critical energy infrastructures by understanding global energy markets
The last day focused on innovation and securing the emerging technologies. The CIO of City of Vienna delivered an insightful presentation about on cities and security implications of digitalisation. A closing panel discussed projected trends and emerging areas of technology, approaches and methods for verifying and securing new technologies and the future of the cyber threat.
What makes a cyber startup successful? From my working with a number of companies, there are four key areas cyber entrepreneurs should consider:
- Are you passionate about the idea?
- How unique is it?
- Can your intellectual properly be protected?
- Do you have genuine expertise in your domain?
- What do people in your community think of you?
- Do you have a strong network and business skills?
- Do you know your client?
- Do you understand their issues?
- Do they trust you to solve them?
- Are you focusing on the right things?
- Are you measuring the right things?
- Are you incorporating client feedback into the development?
The key here, as you can see, is clients. There is really no way around understanding them, pleasing them and focusing on what they want. This feedback will allow you to pivot where required. Above all, stay focused and avoid premature scaling – don’t do too much too soon.
HutZero, an early-stage entrepreneur bootcamp, kindly prepared a list of books and websites recommended for aspiring cyber startup founders.
The Lean Startup, Eric Reis
Business Model Generation, Alexander Osterwalder, Yves Pigneur
The Mom Test, Rob Fitzpatrick
Lean Analytics, Alistair Croll, Benjamin Yoskovitz
To Sell Is Human, Daniel H. Pink
Start with Why, Simon Sinek
The Purple Cow, Seth Godin
Lean UX, Jeff Gothelf, Josh Seiden
Made to Stick, Chip Heath, Dan Heath
The Four Steps to the Epiphany, Steve Blank
Do More Faster: Lessons from TechStars, Brad Feld, David Cohen
Fundraising Field Guide, Carlos Espinal
Wired Threat Level
Krebs On Security
The Next Web
Tech City News
Buffer Blog (marketing)
Fred Wilson’s Blog, A VC
Brad Feld’s Blog (Techstars)
KissMetrics Blog (marketing)
Both Sides of the Table
It’s the second year I’m attending the IoT Security Foundation conference and it continues to be a great event.
Strategic and technical tracks run in parallel with vendor showcases and means that there’s something interesting for everyone.
It’s great to see industry practitioners and academics coming together to discuss the ethics of IoT, challenges with design and development and the direction of travel of security.
Some of recorded talks are available on the IoTSF website.
Best practice guidance on vulnerability disclosure, connected consumer products and security compliance framework are available to download.
A few weeks ago I learnt that my application to attend the HutZero cyber entrepreneur bootcamp had been successful. I am excited to start the programme next week and will keep you posted!
Whether you are just finishing your studies on cyber security, or have worked in the corporate world for a number of years, HutZero supports individuals at the very start of their entrepreneurial journey.