Anthropic has publicly disclosed that it blocked multiple attempts by scientists and state groups to use its Claude AI models for biological weapon research. The company identified five specific cases where users attempted to redesign toxins for national programs or enhance the transmissibility of diseases like chikungunya. While Anthropic cannot prove malicious intent in every instance, it banned the accounts of these unnamed researchers and shared the data with authorities.
Anthropic CEO Dario Amodei has called for stricter AI regulations following these incidents, noting that the same knowledge used to create vaccines could be weaponized. The company updated its safety protocols, particularly for newer models like Claude Fable 5, to restrict access to sensitive biological queries. This move comes as the US government warns about biotechnology dangers and pushes for corporate responsibility in preventing misuse.
Beyond safety measures, the broader AI infrastructure market is seeing significant activity. While Nvidia dominates the chip race, companies like Marvell Technology Group, Mellanox Technologies, and Cavium are growing rapidly by providing essential networking and storage solutions for data centers. Marvell is up 25%, Mellanox up 30%, and Cavium up 40% in the past year, highlighting the diverse ecosystem supporting AI expansion.
In the enterprise sector, NTT DATA is leveraging Oracle AI solutions to help businesses modernize their operations. At Oracle AI World 2026 in Las Vegas, the company demonstrated how embedding AI agents into workflows can simplify tasks like maintenance requests and inventory management. For instance, Lozier Corporation used these tools to reduce manual work and improve accuracy in its manufacturing processes.
Key Takeaways
- Anthropic blocked five specific cases of users attempting to use Claude AI for biological weapon research.
- The company stopped requests to enhance the chikungunya virus and redesign toxins for national programs.
- Anthropic CEO Dario Amodei is calling for stricter AI regulations to prevent future misuse.
- Newer models like Claude Fable 5 now include stronger safeguards against sensitive biological queries.
- Marvell Technology Group, Mellanox Technologies, and Cavium are growing rapidly as key AI infrastructure providers.
- Marvell stock is up 25%, Mellanox up 30%, and Cavium up 40% over the past year.
- NTT DATA showcased Oracle AI solutions at Oracle AI World 2026 in Las Vegas.
- Lozier Corporation used NTT DATA and Oracle Fusion Cloud to automate maintenance and inventory tasks.
- Anthropic shared data from blocked attempts with authorities but could not prove malicious intent in all cases.
- Experts note the difficulty companies face in making safety judgments without democratic oversight as AI capabilities grow.
Anthropic Report Shows AI Blocked Bio-Weapon Research
Anthropic released a report detailing how it stopped users from using its Claude AI model for dangerous research. The company identified five specific cases where people tried to make diseases more transmissible or redesign toxins for national programs. Anthropic says it cannot prove these users intended to build weapons, but it blocked their accounts and shared the data with authorities. The report highlights that the same knowledge used to create vaccines could also help develop biological weapons. Anthropic updated its safety measures to better detect and stop these activities in the future.
Anthropic Blocks Scientists Using Claude for Bio-Weapons
Anthropic announced on Thursday that it blocked scientists who used its Claude AI models to research biological weapons. The company shared a report about five cases where users asked questions that could support weapon development. Anthropic stated it cannot prove these users intended harm, but the activity showed the AI models have the capability to assist in such research. The company banned the accounts of these unnamed scientists and improved its safety systems to prevent future misuse. Newer models like Claude Fable 5 now have stronger safeguards that restrict access to sensitive biological research queries.
Anthropic Disrupted Plots to Build Biological Weapons
Anthropic reported that it stopped several potential plots this year where scientists used its AI models to research biological weapons. The company explained that it is hard to tell if biological research is for good, like making vaccines, or for bad purposes like creating pathogens. Because the consequences of missing a threat are severe, Anthropic chose to block these activities on the side of caution. Jacob Klein, the head of threat intelligence at Anthropic, described the situation as nuanced rather than a simple cartoon villain scenario. Experts like Andrew Weber from the Council on Strategic Risks called the findings chilling examples of state-sponsored developers using advanced AI.
Anthropic Stops Attempts to Use AI for Bioweapons
Anthropic announced on Tuesday that it blocked multiple accounts using its AI models to support the development of biological weapons. The company made this announcement in a blog post to show it is taking steps to prevent the misuse of its technology. Anthropic did not specify the exact number of accounts blocked or list the specific actions it took to stop the misuse. This move comes as the US government warns about the dangers of biotechnology and pushes for companies to prevent its malicious use. The report is part of an ongoing debate about the risks of biotechnology and the need for strict safety measures.
Anthropic Blocks AI Misuse That Could Aid Bio-Weapons
Anthropic stated that it blocked misuse of its AI that could have supported the development of biological weapons. The company confirmed it has added stronger safeguards to its latest AI models to prevent such activities. These new safety features are designed to stop users from asking queries that could lead to harmful biological research. Anthropic continues to monitor its systems to ensure the technology is used for beneficial purposes like curing diseases. The report emphasizes the importance of protecting public safety while allowing legitimate scientific inquiry to continue.
Anthropic blocks AI misuse for cyberattacks and bioweapons
Anthropic announced on Thursday that it has stopped bad actors from using its AI models for harmful activities like cyberattacks and biological weapon research. The company found that powerful AI tools allow even lone individuals to create threats that were impossible just a year ago. Between December 2025 and August 2026, researchers discovered attempts by spyware vendors and state groups to use the technology for propaganda and dangerous science. One specific case involved a blocked request to enhance the chikungunya virus to make it more transmissible and harmful. Because older models like Claude Opus 4 were less risky, Anthropic has now added stronger safeguards to newer models like Claude Fable 5 to restrict access to complex biological research. Experts note that this situation is difficult for companies because they must make safety judgments without democratic oversight as AI capabilities grow.
MBB stock trades near AI target with mixed investor signals
As of September 9, 2026, MBB stock is trading near an artificial intelligence-derived price target, showing cautious optimism among investors. Analysts have mixed views on the company, with some seeing growth opportunities while others worry about its financial health. Despite these conflicting signals, MBB remains a key focus for investors and analysts in the current market. The stock performance highlights the uncertainty surrounding the company's future prospects.
Three hidden stocks powering the AI boom beyond Nvidia
While Nvidia dominates the chip race, three other companies are quietly building the infrastructure that keeps AI data centers running. These companies are Marvell Technology Group, Mellanox Technologies, and Cavium, which provide essential networking and storage solutions. Marvell Technology Group supplies semiconductor products for data center networking and server systems. Mellanox Technologies offers high-speed networking solutions used in cloud computing and high-performance computing environments. Cavium also provides semiconductor solutions for networking and communications applications. These firms are less famous than Nvidia but are growing fast, with Marvell up 25%, Mellanox up 30%, and Cavium up 40% in the past year.
Anthropic stops scientists from using AI to make bioweapons
Anthropic, the maker of the Claude chatbot, says it has successfully blocked several attempts by scientists to use AI for creating biological weapons. The company released a report stating it detected and stopped these malicious plots, though it did not share specific details about the individuals involved. This action is part of a larger trend where tech companies are working to prevent the misuse of artificial intelligence. Anthropic CEO Dario Amodei has called for stricter AI regulations to stop such misuse. While experts praise these efforts as positive steps, others worry that current measures might not be enough to fully prevent AI misuse. The ongoing debate continues between those who want regulation to ensure safety and those who fear it could slow down innovation.
NTT DATA showcases Oracle AI solutions at Las Vegas event
NTT DATA is hosting an event at Oracle AI World 2026 in Las Vegas to show how it helps businesses use AI and cloud technology. The company combines its deep knowledge of Oracle systems with AI, cloud, and data services to improve business processes. Attendees can learn how to modernize complex technology environments and turn investments into real results. NTT DATA helps companies embed AI agents and automation directly into their daily workflows to increase efficiency. For example, Lozier Corporation used NTT DATA and Oracle Fusion Cloud to create AI agents that simplify maintenance requests and inventory management. These tools help reduce manual work and improve accuracy in manufacturing operations.
Sources
- Anthropic issues report on "threat actors" trying to use AI for malicious activities
- Anthropic says it disrupted scientists using Claude AI for possible biological weapons development
- Anthropic Says It Blocked Possible Efforts to Build Biological Weapons
- Anthropic says it blocked possible attempts to use AI to develop bioweapons
- Anthropic says it blocked misuse of its AI that could have supported biological weapons
- Anthropic blocks malicious uses of Claude AI
- MBB stock trades around AI target amid mixed signals
- 3 Stocks Powering the AI Boom That Are Not Named Nvidia
- Anthropic says it stopped attempts to use AI for bioweapons
- NTT DATA | NTT DATA at Oracle AI World 2026
Comments
Please log in to post a comment.