Organized Crime and AI: 5 Topics We Are Monitoring in 2026

As advances in artificial intelligence (AI) bring it to nearly all corners of life, both criminal organizations and law enforcement are adopting these new technologies to try and gain an edge. These are the five areas we are watching most closely in 2026. 

1. AI-Optimized Cyber Crime

AI is making software development more efficient and scalable, and as a result, cybercriminals are finding it easier than ever to develop and deploy custom malware.

Large language models (LLMs), like ChatGPT, now allow people without developed programming skills to write and debug computer code and develop functioning apps with AI’s help. Known in the tech world as “vibe coding,” novice programmers can describe to an LLM what app they want to make, run the code the AI spits out, and then feed the error messages back to the LLM, cycling through this process until a functioning app is produced. 

Vibe coding works well for smaller operations — and crucially, it is very fast — though an inexperienced coder’s lack of programming knowledge can inhibit scaling or the development of more sophisticated apps. Most LLMs have rules prohibiting the AI from doing anything malicious, but they are easily bypassed, and some malware is dangerous not because of what it does but rather how it is used, usually with unwitting victims who click a malicious link or think they are downloading legitimate software.

And when it comes to skilled programmers, AI is making them more efficient. LLMs can help solve tricky programming problems or find vulnerabilities in a target’s security. Perhaps most dangerously, AI “agents,” which, once properly tuned, can launch unending attacks without human supervision — even adapting to security features on its own. Anthropic, the company behind the Claude LLM, disrupted such an AI-driven cyberspying campaign. 

“We believe this is the first documented case of a large-scale cyberattack executed without substantial human intervention,” their report said. And while this case was likely tied to a state-backed group, as AI becomes more powerful and accessible, smaller criminal organizations are likely to embrace its automation for easy profits.

SEE ALSO: 4 Ways AI Is Shaping Organized Crime in Latin America

This increased efficiency is particularly dangerous. Given how many channels cybercriminals have, AI-enhanced attacks are threatening everything connected to the internet, from phones to fridges. Many new smart devices have weak security, making them easier to break into and take over. And while the devices may not be particularly powerful, cybercriminals can chain millions of devices together, known as a botnet, which they use to launch attacks. Or they can hack into the weaker smart device and spread to the computer you paired it to. What’s more, LLMs themselves have now become targets. The interactive nature of these chatbots is quite difficult to secure, and as cybercriminals get increasingly creative, there are already cases of LLMs with access to sensitive information that have been cracked and their data stolen.

2. AI-Optimized ‘Traditional’ Crimes

More traditional criminal organizations have begun to embrace AI to enhance existing frauds and criminal schemes. 

The first example of AI-enhanced fraud in Latin America that came to InSight Crime’s attention was the use of voice cloning for extortion. Groups, particularly in Peru, would take audio from social media and use AI to imitate the individual’s voice, then call that person’s family, pretending they had been kidnapped and needed a ransom. 

But AI-driven frauds became even more sophisticated this year, with groups in Brazil using AI to imitate the faces of victims and bypass biometric security. Technologies once considered the most secure, like face scanning used by governments and banks to secure your accounts, could soon become obsolete thanks to AI. 

And as powerful prison gangs and drug trafficking organizations, such as Mexico’s Sinaloa Cartel and Brazil’s First Capital Command (Primeiro Comando da Capital – PCC) increasingly embrace new technologies, the divide between street and cybercrime is likely to get smaller. With many groups throughout the region recruiting from within prisons, more arrests of cybercriminals may lead to new alliances between traditional and digital criminal groups. 

3. AI-Optimized Drones

Criminal organizations began to use drones years ago for drug trafficking, surveillance, and bombings. But now advances in computer vision are making drone technology an even more effective criminal tool.

The Russian invasion of Ukraine led to small drones, designed for hobbies or filming, being fit with explosives and turned into weapons on the battlefield. This practice is now also being used by criminal organizations in Latin America, perhaps most notably by the Red Command (Comando Vermelho – CV) in Rio de Janeiro, Brazil, where the gang used drones during a deadly raid by police. 

In war-torn Ukraine, AI is now being used to make these “garage drones” more autonomous and accurate, but authorities now worry that these technologies will spread to the underworld as well. “The development of affordable, AI-enabled, and electronic warfare-resistant drones poses new challenges for law enforcement in the context of potential criminal and terrorist abuse,” warned a recent report by EUROPOL. 

Beyond their offensive uses, criminal organizations have long utilized drone technology for surveillance. Mexican drug trafficking organizations, for example, have used drones to spy on rivals and security forces and to intimidate civilians. As AI-driven facial recognition technology becomes more available, criminal groups may combine it with drones to enhance their surveillance capabilities, identifying security officials or anti-crime activists so they can target them or their families. 

4. Governments Look for a New Edge

While criminals are using AI to expand their illicit capacities, governments are trying to strike back, harnessing AI into new and effective crime-fighting tools. AI promises several improvements over humans for fighting organized crime. AI-driven scanners can detect variations in images the human eye cannot see, potentially detecting drug shipments that can currently pass through undetected. AI also does not get tired or bored, making the system less likely to miss an anomaly near the end of a long day. AI is not going to look the other way because someone handed it an envelope full of cash, nor does it have a family that could be threatened if it refuses to let a shipment pass through uninspected.

SEE ALSO: Ghost Riders and Deepfake Doctors: Inside Brazil’s AI-Driven Crime Surge

The applications are potentially endless. In Chile, AI is being used to try to uncover criminal networks. Ecuador announced it would deploy new AI technologies against contraband, illegal mining, and illegal fishing, and the United Nations has called for countries to explore AI as a tool to detect and deter corruption. “Emerging technologies like artificial intelligence have the potential to accelerate corruption but also to help us detect and prevent it,” said Secretary-General António Guterres in a conference in Doha. 

5. The Cure May Be Worse Than the Disease

Despite these advancements, governments — which are beholden to regulations and salaries that cannot compete with the private sector — often lag behind criminal advancements. In trying to keep up, security forces may launch unproved and unrefined technologies at the expense of citizen’s taxes as well as their privacy. 

Inaccurate facial recognition software has already resulted in innocent people being arrested despite evidence that they could not have committed the crime. The software is less accurate when used on people with darker skin, reproducing racial biases, leading to greater errors, and undermining trust in authorities in places like Brazil.     

In Ecuador, the new system is being developed by Palantir, which has received nearly $1 billion in US government funding, raising concerns that it is creating a privatized surveillance state. 

The tools themselves are often not inherently good or bad. They may be used as effective crime-fighting measures or could be used to abuse citizens. As the average democracy score in Latin America has decreased every year for more than a decade, and security is increasingly an electoral priority in the region — even if that means sacrificing rights — tools instituted for the promise of fighting organized crime could easily become tools of oppression.

#border #crime #latinamerica #news