DREAM / The Prime Minister resigns and wakes up under the Parisian bridge!

 Kliko këtu për të tjera
Voxpress:

The dark side of AI: Is it being used for research that could lead to biological weapons?

2026-09-11 17:13:28Kosova&Bota SHKRUAR NGA REDAKSIA VOX

The dark side of AI: Is it being used for research that could lead to biological

US artificial intelligence company Anthropic has announced that it has blocked several attempts to use its models, including Claude, in biological research that could have the potential to develop weapons.

According to a report recently published by the company, several scientists have attempted to use Anthropic's systems for research purposes, while the company has not always been able to determine whether the target was legitimate or posed a risk.

This is precisely where one of the main problems with artificial intelligence lies: the same technology can be used to develop drugs and vaccines, but also to create the most dangerous pathogens.

Jacob Klein, head of threat intelligence at Anthropic, stressed that the cases are not as straightforward as they may seem.

According to him, this is not about people openly declaring that they want to build a biological weapon, but about research that may seemingly have legitimate scientific purposes.

One of the cases that has raised the most concerns involves a scientist who, in May, asked Claude to help draft a funding application for research on the Chikungunya virus.

According to Anthropic, the aim of the research was to induce mutations in the virus to make it more harmful, through repeated infections in animals.

The company also claims that the research would be conducted at a military research facility. Anthropic has not stated that it was a specific attempt to create a biological weapon, but has considered the situation concerning due to the nature of the research and the facility where it would be conducted.

Not just biological weapons

Anthropic's report also includes other instances of its models being used for questionable purposes. The company identified efforts related to propaganda creation, surveillance, and the development of military technologies.

According to the report, Claude has been used or attempted to be used in efforts related to the development of firearms, missiles, armed drones and bombs. The company has also reported cases linked to actors from China, Russia and Yemen.

In another case, the system was reportedly used to create content that posed as independent journalism, but was actually intended to spread propaganda and disinformation regarding the elections in Moldova.

The report comes at a time when the dangers of artificial intelligence are causing increasing debate.

The main concern is that more advanced models may be able to provide more sophisticated assistance in areas where an error or misuse could have serious consequences.

The “thin line” between beneficial and dangerous use

The case of Chikungunya illustrates this dilemma. Research on a virus may aim to develop new vaccines and treatments, but the same knowledge can also be used to make a pathogen more dangerous.

For this reason, Anthropic has decided to intervene and block attempts it considers suspicious, while warning that the capabilities of new AI models are growing rapidly.

The debate has also expanded to social media, where some have raised suspicions that the company's warnings may be part of a communication or marketing strategy.

Others consider these developments a real warning about the dangers that the advancement of artificial intelligence can bring.

Essentially, the Anthropic report brings to the forefront an increasingly urgent question: how far should artificial intelligence go in aiding scientific research, and where should the line be drawn when this technology can be used for potentially dangerous purposes?


Në përputhje me standarded JTI

Video

SHIKO KANALIN NE YOUTUBE
×