In a world accelerating toward automation, comprehensive digitization, and open sourcing, technology transcends the boundaries of peaceful and economic development to open new and concerning doors for armed groups, requiring security vigilance and legislation that keeps pace with the complexities of the modern era and prevents catastrophe.

• Introduction to the democratization of advanced armed violence and artificial intelligence
• Multiple actors and the divergence in current technological capabilities and models
• Terrorists’ advantages and their acceptance of indiscriminate and inaccurate targeting tactics
• The silent threat of unregulated open-source models
• Proactive steps and policies to confront and mitigate impending risks
• Frequently asked questions

Introduction to the democratization of advanced armed violence and artificial intelligence

In an in-depth and concerning analysis of global security shifts, Brian Fishman, co-founder and chief strategy officer at security infrastructure firm Cinder, asserts that advanced violence is undergoing a dangerous democratization process that makes it easily accessible to everyone. Fishman points out that the rapid advancement of artificial intelligence systems, alongside continuous and massive improvements in consumer robotics, complex energy production technologies, and precise navigation and sensor devices, will steadily and increasingly enable small groups of individuals or violent organizations to use targeted violence with greater precision and from safe, distant ranges without immediate consequences. Although security debates often focus on apocalyptic scenarios such as superintelligence taking over the world or rogue states developing advanced and lethal biological weapons—as Dario Amodei warns—the article cautions that empowering these small groups with tactically autonomous tools represents an extremely serious threat that should not be overlooked, as non-state actors will be able to utilize available tools to execute precise operations rather than seeking comprehensive and mass destruction that is difficult to achieve.

Multiple actors and the divergence in current technological capabilities and models

The report clearly indicates that the stunning technological leaps currently witnessed do not restrict their direct benefits solely to government institutions, regular armies, or secret laboratories; rather, they have become accessible to everyone thanks to open models such as DeepSeek and Qwen. Although major tech companies hold the initiative and massive financial capabilities to develop closed and powerful artificial intelligence models used to protect state interests, geopolitical superiority, and cyber defense, this technology conversely provides unprecedented tools for other parties that may seek to spread chaos. Just as the internet revolution in previous decades granted extremist and violent groups broad media platforms to spread their propaganda and recruit followers, the artificial intelligence revolution today offers new and advanced tactical and digital tools that will enable these subversive groups to leverage technology to enhance their capabilities in executing and coordinating field attacks with much greater effectiveness, flexibility, and speed than was available in the recent past.

Terrorists’ advantages and their acceptance of indiscriminate and inaccurate targeting tactics

The fundamental and dangerous problem, as explained by Fishman—who previously served as director of research at a prestigious military academy—lies in the radical difference in operational and ethical standards followed between the parties. States and regular armies strictly require extreme precision and high reliability in autonomous weapons to avoid ethical catastrophes, political fallout, and the destruction of incorrect targets that could lead to civilian casualties. In contrast, terrorist groups demonstrate explicit readiness, acceptance, and a frightening tolerance for very low accuracy standards, focusing clearly on attacking civilian and easily breached targets to instill terror and deliver political messages within society. This notable drop in targeting standards significantly facilitates the adoption and deployment of cheap, low-cost technical solutions, such as integrating simple commercial drones with open-source vision models that can be run locally without requiring an internet connection, and programming them to crudely distinguish between targets based on simple visual indicators such as skin color or clothing type. This is considered completely sufficient and appropriate for achieving their strategic goals and spreading fear, much like the AK-47 rifle did in the past, but in a modern style.

The silent threat of unregulated open-source models

The specialized author strongly warns against the escalating danger posed by open-source artificial intelligence models and open-weight models that are freely available for anyone to use without restriction. Although these software models are considered less complex and advanced than the massive, closed systems developed by well-funded major laboratories, they remain highly effective and capable of carrying out qualitative and dangerous tasks. They can also be easily operated and processed locally on simple computers that do not require continuous connection to monitored corporate servers. More dangerously, these open models can be easily subjected to modification and fine-tuning by extremists to direct them toward malicious and destructive purposes without any oversight, warning, or tracking from developing companies or security agencies. As artificial intelligence continues to lower the technical barrier to entry for programming, individuals lacking high engineering skills will be able to carry out sabotage acts, complex attacks, and develop diseases targeting agriculture and livestock that were previously restricted to advanced military experts.

Proactive steps and policies to confront and mitigate impending risks

The expert offers policy, legal, and practical recommendations to deal firmly with this changing and complex reality, emphasizing first and foremost that legislative authorities and councils must immediately initiate updates to federal criminal laws to classify any deliberate modification of smart models aimed at facilitating and directing violence as a criminal act falling under premeditated planning, considering it material support and funding for terrorism punishable by extremely severe laws. Furthermore, the article calls for the necessity of adopting practical and field defense strategies, such as fortifying and securing potential targets, deploying advanced jamming systems, and utilizing specialized electronic countermeasures to shoot down cheap drones before they reach their targets. It also strongly urges developing companies and those hosting open models to build and form professional, specialized analytical teams to track suspicious usage patterns, share information, and limit misuse as much as possible, similar to the efforts made in major foundational model laboratories that monitor activities to prevent biological and security disasters from occurring.

Frequently asked questions

Question: What is the main tactical threat highlighted by the article regarding artificial intelligence?

Answer: The threat lies in the potential for artificial intelligence to enable small and terrorist groups to use autonomous, precise, and directed violence to carry out armed attacks remotely via drones.

Question: Why is it easy for armed groups to adopt autonomous weapons despite their rudimentary and simple nature?

Answer: Because these groups accept very low accuracy standards and are willing to attack random civilian targets, making it easy for them to use and direct low-cost drones for their missions.

Question: What is the security and legal danger of open-source artificial intelligence models?

Answer: Their danger lies in the ease of running them locally on simple devices, and the possibility of modifying and programming them to carry out malicious tasks without any restrictions or direct, centralized oversight from companies.

Question: What legal and legislative recommendations are strongly proposed to combat these emerging threats?

Answer: The expert recommends that legislators update laws to classify the deliberate modification of models to facilitate violence as an explicit crime and direct material support to extremist terrorist organizations.