Soon, it is going to be AI Vs AI battle in cybersecurity, fake news and defence

Deepak Garg
4 min readJul 3, 2021


Applications of Artificial Intelligence technologies are showing their impact on society and are growing very fast. It is revolutionizing all industry verticals and social dimensions. With most AI tools available as open sources and the cheap availability of cloud computing resources, it is possible for anyone to try their hands on such technology.

A lot of debate is going on regarding the ethics, morality and the legality of the potential applications that can be built using AI. Powerful non-state actors who do not care about the intended consequences of technology for their short-term gains in terms of money or power- make this accessibility seem baleful. Even many governments support skilful dubious actors that may be used against the adversary without opening a formal war front.

This trend has been clearly demonstrated — in the context of cyber hacks. In cybersecurity, it is always a race between hackers and protectors, and as a result of this heated contest, businesses are making money. Cybersecurity Companies leave no option for organisations, but to invest in robust security to avoid potential cyber-attacks. With the advent of AI, both sides are now making sophisticated systems to out-do each other. We have recently seen AI applications where trained models can detect the signatures of your network traffic to classify it as a genuine or fraud and based on that you may act in real-time to avoid these attacks. Similarly, attackers are also trying to hoodwink the AI systems by changing the attack patterns which may not have been learnt by AI models. So, it is fast developing as AI Vs AI in cybersecurity. Concurrently, AI systems are also becoming smart to guess potential passwords and break the security of individuals. In the end, what would matter is which one would win in this seemingly (and possibly) never-ending technological war.

Similar situations are emerging in the AI space, however now with much higher stakes. We will be deliberating fake news and the defence sector, which are two most important flashpoints in this case.

Fake news

Fake news is emerging as a topmost concern for society. It has the potential to trigger riots, anarchy, disillusion, and chaos. Currently, Planned targeting of common people with manufactured videos, articles and messages has become a common tool for political, social and terrorist groups to set their agenda and increase their followers. In the long term, it has the potential to divide society into different groups having extremist and polarizing views about each other.

With the advances in Generative Adversarial Networks and other AI techniques, the generation of fake images, fake audio and videos is becoming an easier task by the day. It can be done by anyone with minimal resources.

The government, media organisations and social media companies are driving efforts to detect fakery and hence remove fake news using AI. The overarching result of the kind of information available to us is dependent on the development of the generation of fake artifacts verses the detection of these fake artifacts. Which side will succeed and create better AI is uncertain, however, it is certain that the public will face hardships due to the fake news- all the while companies compete their way into gaining more money. So eventually it is a race between AI and AI.


Trillions of dollars are being spent to augment the defence capabilities using AI and other trillions to safeguard the assets from AI enabled warfare by adversaries. AI capable nations will become a better tag than a nuclear capable nation in the coming decade. With AI you can choose where to attack, how much loss to cause to the adversary and deploying tactics — without starting a real war.

Countries are spending a lot of money in AI to enhance their surveillance, reconnaissance and intelligence capabilities. Tremendous amounts of public resources are being consumed to make AI-enabled autonomous war vessels, miniature submarines, swarm drones, smart missiles and ammunition, robot soldiers, micro kill machines. Even if governments are starting a new AI arms race the poor public must accept the logic that it is being done for their safety. Whether these new defence technologies are being developed to attack or defend, either way taxpayers’ money is being floundered which may have been spent in fast-tracking the sustainable development goals of humanity. Once again — AI faces AI in the race of making sophisticated defence technologies.

There are many other areas where first AI will be developed to increase the efficiency, productivity and quality of life and then another set of AI will follow to disrupt such systems. The race will go on.



Deepak Garg

Professor and Chair, CSE, Bennett University; Director NVIDIA Bennett Research Center on AI; Director