Terrorism has been a major issue in Pakistan, and the country has been fighting many terrorist organizations that have threatened national security. Terrorism has been changing in nature, and it requires a new approach to eliminating the menace. Artificial Intelligence has been a critical new approach to counterterrorism since it provides state-of-the-art solutions that enhance national security. With AI technology, Pakistan can change its fight against terrorism, such as intelligence collection, threat identification, and response time.
Artificial intelligence involves the study of how digital computers or machines can perform activities similar to those done by intelligent creatures. Such behaviors include thinking and planning, discovering, and making, understanding and interacting and embodied on efficient processes, knowledge representation, and extrapolating existing digital data. In addition, AI targets activities that require intricate mental operations and processing. Counterterrorism experts suggest that there exist two major ways to prevent terrorist activities: first is through providing safety to people and vital infrastructure conditions to implement security. Secondly, criminals have the capability to get hold of terrorists before they commit operations. The last option impedes extremists from recruiting terrorists.
Artificial Intelligence (AI) encompasses the capability of digital machines and computers to execute tasks in a manner analogous to that of intelligent organisms. These tasks may include but are not limited to reasoning and planning, learning, and generating, adapting and interacting, enhancing procedures, extracting knowledge, and forecasting vast and diverse digital data. Furthermore, AI extends to other operations that demand precise mental processes.
Experts in counterterrorism define two approaches to averting terrorist attacks: safeguarding infrastructure and individuals while implementing security controls is the initial method; depriving terrorists of the ability to execute their plans by apprehending them prior to their execution and combating extremism and terrorist recruitment constitutes the second method.
The article by Anna Rosner, Alexander Gegov, Djamila Ouelhadj, and Adrian Alan Hopgood titled “Neural network-based prediction of terrorist attacks using explainable artificial intelligence” examines the use of artificial intelligence (AI) to forecast terrorist attacks. The advent of AI has revolutionised the domain of terrorism prediction, enabling law enforcement agencies to detect prospective threats with significantly enhanced velocity and precision. The authors of the article put forth an innovative utilisation of a neural network in order to forecast the “success” of a terrorist attack. An F1 score of 0.954 and an accuracy of 91.66% are achieved by the neural network. The obtained accuracy and F1 score surpass those of alternative benchmark models. However, there are limitations to using AI to anticipate high-stakes decisions, such as potential biases and ethical concerns.
According to leaked data from the US National Security Agency (SKYNET) programme, in 2007, an AI-based algorithm was used to analyse the metadata of approximately 55 million mobile phone users in Pakistan, putting approximately 15,000 out of a population of 200 million at risk of becoming terrorists. While the model employed failed to achieve true effectiveness, it did demonstrate the predictive capability of the data in detecting intimate associations with terrorism. While the applications of predictive AI in counterterrorism remain viable prospects, it is not reasonable to expect AI to furnish instantaneous, exhaustive, and precise resolutions to intricate inquiries. Using AI, Pakistani intelligence agencies can identify terrorists and implement preventative measures against future attacks.
Furthermore, Pakistani intelligence agencies can use artificial intelligence to predict the date and location of terrorist attacks by evaluating communication data, financial transaction information, travel patterns, and internet surfing activities. Predictive models regarding the location and timing of terrorist strikes have been developed. In 2015, for example, a technology startup claimed that its prediction model was 72% accurate in forecasting suicide strikes.
A number of additional models, such as a preemptive event recognition system that integrates the outcomes of distinct predictive models to forecast particular events, have also utilised open-source data pertaining to mobile phone users and those who utilise social media. It is imperative to recognise that an increase in data does not inherently imply an improvement in the quality of the prediction; instead, it is necessary to validate the assertion.
Let’s analyze the data of 2024 attacks and predict the future attacks by these proscribed organizations:
| Attacks | Location | Casualties | Terrorist Organization |
| Suicide Bombing in Karachi | Karachi, Pakistan | 3 bystanders injured, 1 accomplice of the suicide bomber | Small separatist groups or Pakistani Taliban |
| Terrorist Attack Against Chinese Nationals | Besham, Khyber Pakhtunkhwa province, Pakistan | 5 Chinese nationals and 1 Pakistani national killed | Enemies of Pakistan-China friendship |
| Balochistan Bombings | Pishin District and Killa Saifullah, Balochistan, Pakistan | At least 30 people killed, at least 40 people injured | Islamic State -Khorasan Province |
Now, let’s predict future attacks and their locations. It’s evident that AI has the capability to understand patterns and make predictions. However, with limited and uncomplicated data, using complex algorithms and considering intricate situations, AI can perform better.
| Region | Attack Patterns | Potential Targets for Future Attacks |
| Khyber Pakhtunkhwa (KPK) | Attacks on police stations, military outposts, and security forces | Peshawar: Major urban center and vulnerable to attacks on government buildings and police stations. North Waziristan: Near the Afghan border, susceptible to cross-border incidents and attacks on security personnel. |
| Balochistan Province | Insurgent activity, attacks on security forces, and infrastructure sabotage | Quetta: Provincial capital and economic hub, potential targets include government buildings, security checkpoints, or public places. Gwadar: Strategic port city, vulnerable to attacks on infrastructure or foreign interests. Panjgur: Previously targeted, potential for future attacks on security forces or infrastructure. |
| Punjab Province | Occasional attacks on intelligence officers and government installations | Lahore: Major city could be targeted for attacks on government buildings or security personnel. Mianwali: Previous attack on a police station, potential for similar incidents. |
Moreover, many technology businesses have developed algorithms for determining sensibilities to violent extremist conducements. One used for such a project, aimed at the consumers of video sharing platforms who might be impressionable to terrorist propaganda and redirect them to videos that promote a plausible counter-narrative, was established by this company.
Finally, Pakistan can implement artificial intelligence technology to curb terrorism by foretelling the duration and place of terrorist attacks. This can be done by primarily analyzing communication data, information on financial transactions, travel mediums, and deeds related to internet surfing. The use of artificial intelligence allows Pakistani intelligent agencies to detect terrorist cells and implements preventative measures to avert prospective attacks.
Be the first to comment