To mark its 15th anniversary, the European Union Agency for Cybersecurity (ENISA) published the seventh edition of its Threat Landscape report.
Looking at the changes from 2012, we see a range of threats that we haven’t seen previously and old threats with increasing sophistication and destructive power.
Adversaries, modus operandi and motives are changing over time, with hackers becoming entrepreneurs by setting up businesses or even working for governments as cyber agents.
These changes have a profound impact on the importance of cybersecurity, as it has become possible to use cyberspace to shape elections or shutdown a country’s power grid.
Cybersecurity is assuming an increasingly critical role in economy and society, highlighting our growing dependency on cyberspace to run our daily lives.
There are numerous reasons for this rate of change. However, one reason stands out; a pattern that is changing the cybersecurity paradigm - the ‘datafication’ of everything
This is a pattern emerging from the quest to transform every aspect of our human existence into quantified data.
From a simple perspective, this ‘datafication’ is about recording any type of activity and turning it into data that can be monitored, tracked, analysed and optimised. What makes ‘datafication’ a pattern is mainly the fact that everyone loves data.
Governments use it to define policies, plan strategies and allocate resources; researchers use it to stimulate advances in science; industries use it to monitor and control manufacturing processes and businesses to study and define strategies that attract and influence customers.
As individuals, we are also addicted to data, behaving like ‘dataholics’. The amount of data collected, stored and processed per person is growing as we move more information and activities into the digital space.
For example, the use of visual data such as images, pictures and videos will become a norm for individuals to communicate, replacing the existing text-based approach.
Audio data is already growing beyond traditional telephony systems, with digital personal assistants gaining growing importance in our digital lives.
“Cybersecurity is assuming an increasingly critical role in economy and society, highlighting our growing dependency on cyberspace to run our daily lives”
Cybersecurity - and threat analysis in particular - relies on data, defined as indicators of compromise.
The deluge of data generated by security monitoring and detection systems goes beyond what is humanly possible to handle and digest.
With advances in data analytics and artificial intelligence (AI) tools, the processing and analysis of all this data may finally become possible.
The insights produced by AI algorithms can be used to train defence systems to protect themselves from cyberattacks.
This perspective holds great promise; however adversaries can also leverage the potential of AI to breach traditional defences and undertake attacks.
One of the threats relates to the fact that AI algorithms are a black box, where audit and validation are simply not possible.
As an example, any malicious modification of the data used to train AI algorithms for autonomous systems can have unforeseeable and devastating consequences on how a vehicle behaves.
Other threats include the use of AI to automate tasks involved in surveillance - such as analysing bulk-collected data; persuasion - through creating targeted propaganda; and deception - manipulating videos and facial recognition.
Novel attacks may also take advantage of an improved capacity to analyse human behaviour, moods and beliefs on the basis of available data.
“New strategies will reinforce protection mechanisms, but it will remain the responsibility of individuals and organisations to adopt stringent criteria on what to share, collect and retain”
For all these, it is reasonable to assume that the appetite for, and the importance of, data will only escalate. This makes its protection of paramount importance.
‘Datafication’ is changing the nature of cybersecurity and how we view threats and protect systems from potential attacks.
The danger of inappropriate use of data goes beyond privacy and data protection issues and expands beyond traditional threats.
It includes how data, collected for valid and legitimate purposes, may be unethically and sometimes illicitly used - through the use of AI algorithms - to limit, sway and manipulate individuals’ actions, opinions and decisions.
The boundaries of the acceptable collection, transmission, processing and use of data need to be clearly-defined to protect against the risk of abuse.
In future editions of ENISA threat landscape, we will see new trends associated with the abuse of data from the “datafication”.
This ‘weaponisation’ of data by using AI will be seen as an alarming trend, likely to be a leading future threat.
These concerns will be further exacerbated by the fact that it will take time for cybersecurity professionals to identify and mitigate these type of threats and to reduce their impact.
New strategies will reinforce protection mechanisms, but it will remain the responsibility of individuals and organisations to adopt stringent criteria on what to share, collect and retain.