Anthropic Says It Blocked Possible Efforts to Build Biological Weapons
Anthropic says it has disrupted several cases this year in which scientists used Claude for biological research that could potentially aid bioweapons development. "In a report describing misuses of its A.I. models, Anthropic said it could not determine whether the research served a legitimate or nefarious purpose because valid biological inquiry -- the kind that can lead to breakthroughs like…
Anthropic has thwarted multiple instances in which researchers have utilized its AI model Claude for biological research that could potentially contribute to the creation of biological weapons, according to a recent report by The New York Times. The tech company explained that it was unable to ascertain whether the research was intended for legitimate or malicious purposes, as legitimate scientific inquiry capable of leading to advancements such as vaccines could also aid in the development of dangerous pathogens.
Anthropic chose to adopt a cautious approach due to the grave implications of missing malicious activities, as the consequences of overlooking such threats could be severe. Experts express deep concern over the potential of advanced AI models to facilitate the creation of biological pathogens, a concern that is becoming increasingly relevant given the rapid advancement of AI technology.
Andrew Weber, a senior fellow at the Council on Strategic Risks, deemed the findings "chilling examples of state-sponsored biological weapons developers leveraging the rapidly evolving capabilities of leading AI models." The comprehensive report, published by Anthropic on Thursday, detailed a range of unauthorized uses of its AI models, including Claude, over the past eight months.
These instances mirrored prior disclosures from Anthropic and other AI labs, such as suspected government-linked actors from China and Iran targeting dissident and diaspora communities for surveillance. Additionally, the report highlighted Russian state media employing Claude to produce online propaganda with fabricated claims about an election in Moldova.
A new form of misuse identified by Anthropic involved attempts to leverage Claude for developing software for traditional weapon design and deployment, including firearms, missiles, armed drones, and bombs. The report documented such cases in China, Russia, and Yemen, although the specific party involved in the Yemeni case remains unnamed, with the context indicating it pertains to the Iran-backed Houthi militia.
Written by urgent.news from Slashdot's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
This story
This is one outlet's version. Read the fullest account.
- Anthropic says it blocked misuse of its AI that could have supported biological weapons toronto.citynews.ca
- Anthropic says it blocked possible efforts to create biological weapons cbsnews.com
- Anthropic Says It Blocked Possible Efforts to Build Biological Weapons nytimes.com
- Anthropic Says It Blocked Possible Biological Weapons Undertaking forbes.com
- Anthropic says it blocked misuse of its AI that could have supported biological weapons cnbc.com
- Anthropic says it blocked attempts to use AI for potential biological weapons ft.com