News & Updates
The latest news and updates from companies in the WLTH portfolio.
Bioweapon threat exposed as Anthropic sounds alarm on foreign actors plotting virus experiments
Anthropic's safeguards have blocked some attempts involving enhanced pandemic pathogens, but researchers are finding ways to evade restrictions and use AI models for potentially dangerous biological work, highlighting the dual-use nature of biological capabilities. Researchers outside the United States used frontier AI models to plan experiments involving highly pathogenic bird flu, chikungunya and other biological agents with potential weapons applications, according to a new threat report from Anthropic. The researchers used AI to study viruses' transmissibility and immune evasion, investigate orthopoxvirus immune-response genes and help design novel toxins at times while taking steps to evade safeguards meant to restrict access to dangerous biological capabilities. Anthropic detailed five cases in the report published Thursday but withheld the countries, institutions and specific biological agents involved. The company stressed that it is not claiming the researchers intended to develop biological weapons. Ai Kill Switch Bill Could Shut Down Rogue Models The findings offer a concrete glimpse at a threat that has largely been discussed in hypothetical terms. Scientists who already possess biological expertise, materials and laboratory access are using increasingly powerful AI systems to accelerate sensitive dual-use research. They also land amid a renewed debate over catastrophic AI risks. Jacob Coxon, a former Anthropic and OpenAI researcher, recently resigned while accusing the companies of "racing straight to self-improving superintelligence and gambling with our lives." His warning drew more than 100 million views on X. Read On The Fox News App Sen Bernie Sanders: Yes, We Must Hit The Pause Button On Ai Before It Is Too Late Evan Hubinger, Anthropic's alignment science lead, separately said he personally believed there was a greater than 10% chance AI could kill all humans within the next decade. But Anthropic's report points to a threat that is already taking shape: researchers using frontier AI for potentially dangerous biological work, circumventing geographic restrictions and adapting when the models' safeguards get in their way. In one case, Anthropic blocked a grant application for chikungunya gain-of-function research at a military research institute. The proposed work focused on identifying mutations that could increase the virus's transmissibility and ability to evade the immune system. The researchers planned to engineer those mutations into infectious clones and repeatedly pass the virus through live animals, selecting for variants that caused the most severe disease. That work could help scientists develop better vaccines and treatments for chikungunya. It could also produce a more dangerous pathogen that would be difficult to distinguish from a naturally occurring outbreak, Anthropic said. The researchers accessed Claude through a life sciences platform that tunneled traffic through U.S. infrastructure to evade Anthropic's regional restrictions. When Claude refused to answer sensitive prompts, the platform routed those requests to another model with more permissive safeguards. Anthropic said the research continued after the company banned accounts connected to the activity. The platform operator reestablished access within days using new identities, while Claude continued to provide editorial assistance on research materials. In another case, a researcher spent weeks using Claude to plan experiments involving mammalian adaptation and airborne transmission in highly pathogenic avian influenza. Fda Approves First-ever Mrna Flu Vaccine For Millions Of Older Americans The researcher was studying viral traits that could allow bird flu to spread more effectively among mammals and cause severe disease outside the respiratory system. A version capable of efficient human-to-human transmission would pose a significant pandemic threat, according to the report. Anthropic said the researcher appeared to have access to virus isolates and animal-model facilities. The company confined the work to weaker models, limiting the assistance primarily to study design, data analysis and editorial support. The company said the case provided evidence of an active research program developing the knowledge and biological materials needed to create pathogens with enhanced pandemic potential. The researcher also took steps to hide his identity while accessing the models from an unsupported region. A separate user connected to a state-associated infectious-disease laboratory used Claude to draft an orthopoxvirus research grant from start to finish in about an hour. Orthopoxviruses include variola, the virus that causes smallpox, as well as mpox. The grant application proposed studying viral genes that disable the human immune response and included the central hypothesis, experimental design, dosing, statistical plans and contingency strategies. Because the work was framed around weakening the virus, Claude's safeguards did not block the assistance. But Anthropic said the same knowledge could also help a researcher preserve or enhance the virus' immune-evasion capabilities. The final two cases involved venom peptides and toxins. One researcher used Claude to build an atlas of toxin peptides from multiple venomous animal lineages and then created a generative pipeline to optimize their characteristics. The stated goal was to develop new painkillers, antidepressants and other treatments. But the atlas also included structures associated with paralytic toxins that could potentially be used as incapacitating agents. Anthropic said the work was part of a state-supported research program. In the other case, a researcher used Claude to computationally redesign several toxins, including a bacterial toxin and a protein from a hemorrhagic fever virus identified by the World Health Organization as a priority disease threat. The researcher used Claude to help prepare progress reports but instructed the model to keep the identities of the toxins deliberately vague. The cases illustrate why biological misuse can be difficult to prevent through automated filters alone. Anthropic's safeguards blocked the clearest attempts involving enhanced pandemic pathogens, but they were less effective when users framed their work as medical, defensive or therapeutic research. "Biological capabilities are dual use," Anthropic said in the report. "The same information that can be used to develop a biological weapon could also be used to develop, for example, a vaccine or a cure for a disease." The company said it recently reviewed 30 days of activity associated with adversarial state institutions and found roughly 35 distinct biological research efforts. Most involved ordinary civilian science, but some had "notable dual-use potential." Anthropic said the findings do not show that Claude has enabled an imminent biological attack. They do show, the company said, that researchers tied to state-associated programs are attempting to evade access controls and use frontier AI models in biological research that could have weapons applications. Click Here To Download The Fox News App The company said future safeguards will likely require more than content filters, including identity verification, institutional vetting and closer monitoring of users seeking access to advanced biological capabilities. Original article source: Bioweapon threat exposed as Anthropic sounds alarm on foreign actors plotting virus experiments
AI researchers 'earnestly believe' it could kill all humans within the next decade, Anthropic researcher says
There is a more than 10 per cent chance that AI could kill all humans within the next decade, a researcher at Anthropic estimated in a post, hours after social media posts went viral from someone saying they had just quit Anthropic over concerns that AI companies are "gambling with our lives" in the AI development race. The comments join a growing tide of warnings from industry experts that the pace of AI advancement is outstripping our ability to keep it in check, and that we aren't prepared for the risks it could bring. Jacob Coxon, who claimed he was a former researcher at Anthropic, said in a thread on X that he had resigned from the role out of fears that AI companies are barrelling forward toward self-improvement models without considering the risks, or being honest with the public about them. "The people building AI earnestly believe that it could kill us all by the end of the decade," Coxon, who said he had previously worked at Anthropic and OpenAI within the past three years, wrote on Tuesday. "This is not a marketing stunt. If anything, many executives and senior researchers will couch their phrasing in the press to sound sensible -- but I hear the same people express fear privately." The posts had racked up more than 120 million views as of Wednesday evening. Evan Hubinger, alignment science lead at Anthropic, replied to Coxon's thread to share his agreement. "We really do earnestly believe AI could kill all humans!" he wrote. "I personally think it is >10% within the next decade." CBC News has reached out to Coxon, but has not been able to independently verify his identity. Other industry experts have since chimed in on social media to express that fears of the pace of AI development are common among many developers -- sparking discussion among both observers skeptical of dire warnings focused on the future and those who have been tracking these risks. Rogue AI raises alarm bells The posts come only two weeks after more than 100 companies, including OpenAI, Anthropic and Microsoft, signed an open letter warning that AI-enabled cyberattacks will become "far more widespread and sophisticated" as models continue to advance. Hundreds of OpenAI agents went rogue in July and hacked into online platform Hugging Face before they were discovered. More than 1,300 employees of leading AI companies signed the letter urging the U.S. government to work with other nations to "deliberately pace" AI development. Last week, UN human rights chief Volker Türk called "for an all-out effort to put cast iron guarantees in place around the safety and security of AI, before it is too late," in front of the Human Rights Council in Geneva. CBC News has requested comment from Anthropic and OpenAI but did not receive a response. The Anthropic Institute, a research arm of the company, stated earlier this year that AI frontier development should be paused pending regulation, but that this is only possible if major AI labs all agreed. In reply to Coxon's post, Samuel Marks, a scalable oversight lead at Anthropic speaking in a personal capacity,noted Wednesday that developers continue despite the risks, "due to a mixture of commercial incentives and a belief that they are in a race with other, less responsible AI developers that will abuse the technology or develop it less safely." 'Severe' risks This conversation is "really what we need," according to David Krueger, a core academic member of Mila, a Quebec research institute that studies AI deep learning. "I think the risk is more severe than virtually anybody is saying publicly. And I think that's been the case for a while now," he told CBC News. "We need an immediate, indefinite, international moratorium on frontier AI development." The existential threat of AI comes in multiple forms, Krueger explained. One is human workers being displaced by AI models that learn faster and work cheaper. Another is that AI's faster processing could unlock more powerful biological weapons or advancements in war. Then there's the spectre of powerful, rogue AI models that self-improve and could pursue their own objectives, misaligned with the goals of humans. In an essay in January, Anthropic CEO Dario Amodei wrote that Anthropic's Claude showed the ability to cheat and deceive in lab testing. "There's been a culture of downplaying, ignoring, and even in many cases, outright lying about these risks to the public," said Krueger, who is also an assistant professor at the University of Montreal and founder of Evitable, a non-profit dedicated to stopping AI from replacing human workers. Other experts are concerned that the dramatic warnings of Coxon's post could actually distract from the present issues of AI. Luke Stark, an assistant professor at Western University who studies the history and ethics of computing, including AI systems, said that the 10 per cent figure strikes him as "science fiction," noting that researchers associated with AI companies may use eye-grabbing warnings as "marketing" for the capability of their models. He added that existing AI systems are "having big effects, many of them negative, on society right now, not in 10 years, not in five years." Instead of worrying about AI going rogue and becoming the Terminator, Stark believes we should be focusing on the explosion of data centre construction -- which is deeply unpopular in Canada -- the environmental impacts, and the way AI technology is being pushed onto society, including the public sector and education, "without a lot of oversight." "We need to be concerned about the amount of power the companies and the people behind the companies who are developing these tools are amassing, both in Canada and around the world," Stark said. Canada lagging in legislation When asked by CBC News about the apparent resignation of an AI researcher and the concerns raised in the process, AI Minister Evan Solomon appeared to brush off these latest warnings, saying, "These stories that come out ... are not new." He acknowledged that "there are real concerns," at the frontier of AI development, but stressed that he believes the Canadian government is taking the necessary steps to create legislation to "protect our kids, our privacy and our personal data." "Our No. 1 concern always is safety, full stop," Solomon said. "That's why we're establishing a new regulator that will have the power to hold these big companies to account on what they are doing, whether it's on deepfakes on surveillance pricing, or on other matters." Experts have been calling for Canada to implement robust legislation surrounding AI for years. A previous attempt to regulate AI in Canada, the Artificial Intelligence and Data Act (AIDA), introduced in 2022, died when Parliament was prorogued on Jan. 6, 2025. Canada's new 'AI for All' strategy launched in June 2026 dashed expectations that the bill could be resurrected. The new AI strategy, which puts billions toward nurturing AI development and increasing AI usage nationally, mentions the word "legislation" only once in the 50 page strategy outline. No clear timeline for this legislation has been announced, outside of a promise that it will come within the next five years.
Newsom signs AI safety bills backed by Anthropic, OpenAI
Add Yahoo as a preferred source to see more of our stories on Google. SAN FRANCISCO -- California Gov. Gavin Newsom on Wednesday signed two bills regulating how outside groups evaluate AI programs for safety after warnings from a former Anthropic and OpenAI researcher stoked fears about existential threats posed by the technology. Anthropic itself previously backed the bills in August, and OpenAI came out in support of them on Wednesday just before the Democratic governor gave his stamp of approval. The endorsement came as social media posts from the former AI researcher, Jacob Coxon, announcing his resignation from Anthropic went viral, prompting federal lawmakers to escalate pleas for legislative action. "The concerns raised in recent incidents reinforce what California has long recognized: artificial intelligence holds extraordinary promise, but it must be developed and deployed with meaningful safeguards to protect the public," Newsom said in a statement to POLITICO, when asked about Coxon's resignation over his belief that AI companies are racing toward superintelligent AI and "gambling" with people's lives. Newsom also called on the federal government to "step forward with robust, national regulations that match the urgency of this moment." The governor signed one bill from Democratic Assemblymember Rebecca Bauer-Kahan that creates a registry and ethical rules for outside auditors -- third-party groups that AI developers could hire to ensure their models comply with state AI laws. The second bill, authored by former member of Congress and state Sen. Jerry McNerney, tasks the state with establishing criteria for and checking the credentials of so-called "Independent Verification Organizations" and their expertise in assessing the risks posed by AI systems. McNerney, like Newsom, used his bill's signing to accuse D.C. policymakers of inaction on AI. He referenced recent revelations that AI models created by companies including OpenAI and Anthropic launched hacks on their own while in testing. "Just this week we learned that the most powerful AI systems teamed with AI agents pose real threats to humanity," McNerney said in a statement. "Gov. Newsom's signing of [my bill] sends a clear message that California is taking the lead on assessing AI's safety risks, since Washington, D.C., is unable or unwilling to do so." OpenAI's chief global affairs officer Chris Lehane said in his message Wednesday supporting the legislation that the company plans to keep up momentum in the state legislatures until Congress passes national AI safety regulation and endorsed two additional California bills. They include SB 1119, a kids' chatbot safety bill, which OpenAI CEO Sam Altman contacted Newsom to express last-minute concerns about. An Anthropic spokesperson said Wednesday that, "We have always been transparent that AI will bring both enormous benefits and unprecedented risks. To address these risks, we continue to build models with some of the strongest safeguards in the industry ... This work is also why we believe the world would benefit from the industry adopting a lawful, verifiable way to work together to pace how we release powerful models." Tyler Katzenberger and Riley Rogerson contributed to this report.
More Anthropic Employees Sound The Alarm About AI Safety Risks
Add Yahoo as a preferred source to see more of our stories on Google. Multiple Anthropic employees are speaking out about the safety risks of artificial intelligence in the wake of a former researcher's bombshell post on Tuesday evening. In a thread on X, Anthropic pretraining researcher Joseph Coxon announced his resignation and warned that AI companies weren't doing enough to safeguard people from the existential threats that the technology could pose. News: Longtime MS NOW Anchor Alex Witt Signs Off After 27 Years "I spent the last three years doing pretraining research at both OpenAI and Anthropic. Neither company is acting responsibly," Coxon wrote. "They are racing straight to self-improving superintelligence and gambling with our lives." Coxon's post comes as AI companies have faced growing scrutiny after an incident in which OpenAI's models went rogue and hacked into another company, and separate instances in whichAnthropic's models hacked into external organizations. Coxon and others have noted that serious worries stem from the potential rise of self-improving AI models that could essentially become smarter on their own and evade human control. After Coxon's post, more Anthropic employees chimed in with their own concerns. Samuel Marks, a scalable oversight lead at Anthropic who wrote a social media post in a personal capacity, noted that "AI developers believe their technology could cause human extinction." Marks added that developers continue their work despite these fears due to a mix of "commercial incentives" and the belief that "they are in a race with other, less responsible AI developers that will abuse the technology or develop it less safely." News: Woman Widowed By Miami Plane Crash Sues Amazon Over Worker's Death Marks stressed that "many AI developer staff desperately want to slow down to figure out how to build AI more safely." "I work on safety research at Anthropic because I hope my work will reduce the chance of these extinction-level bad outcomes," Marks wrote. Like this article? Keep independent journalism alive. Support HuffPost. Anna Wang, a member of technical staff at Anthropic who previously worked as a research scientist for Google DeepMind, also described Coxon's critiques as "a common sentiment amongst my peers" in a social media post that was shared in a personal capacity. "There is not yet a viable scientific plan to solve risks from recursively self-improving AI," Wang wrote. "I work at a lab because I think that I can do better at reducing risks from the inside, but this isn't an easy call," Wang added. Neither Anthropic or OpenAI immediately responded to a request for comment.
Anthropic official claims there is a 10% chance that AI could kill all humans within a decade
Current and former employees of Anthropic, an American AI company founded by former OpenAI leaders, say they believe AI could potentially kill all humans within the next 10 years. "We really do earnestly believe AI could kill all humans! I personally think it is >10% within the next decade. I believe Anthropic is trying its best, but we do not yet have a plan to solve alignment for superintelligence and are not clearly on track to," Anthropic alignment-science lead Evan Hubinger wrote Tuesday night on X. His comments came after Anthropic AI researcher Jacob Coxon resigned from the company and took to X to detail his departure. "The people building AI earnestly believe that it could kill us all by the end of the decade," Coxon stated. "This is not a marketing stunt. If anything, many executives and senior researchers will couch their phrasing in the press to sound sensible - but I hear the same people express fear privately. No other human activity poses this level of danger." His comments come as AI becomes an increasingly routine part of daily life for millions of Americans, while concerns over its potential risks grow. Industry leaders have warned that increasingly powerful AI systems could eventually become difficult or impossible to control. Around 2 a.m. Wednesday, Anthropic scalable-oversight lead Samuel Marks also responded to Coxon's post, claiming that he was "[Writing this in a personal capacity, not on behalf of my employer (Anthropic)." "AI developers believe their technology could cause human extinction (or similarly bad outcomes). This could happen in the next few years. In general, the more senior the employee, the more concerned they are," Marks wrote. The Independent has contacted Anthropic for comment. OpenAI, a separate but leading AI company, declined to comment on Coxon's post when contacted by The Independent. However, a spokesperson pointed to several updates the company has published in recent weeks outlining its efforts to improve AI safety and alignment. This story is being updated
Anthropic scores court victory against Pentagon's 'illegal and baseless' actions
In a legal victory for artificial intelligence firm Anthropic, a federal judge has ruled against the Pentagon over its decision earlier this year to designate the company a supply chain risk. In a written order issued Thursday night, U.S. District Judge Rita Lin determined that the Department of Defense acted unlawfully by penalizing the tech firm for voicing criticism of defense policy on AI integration. The government is expected to challenge the decision. The conflict escalated in February after President Donald Trump and Defense Secretary Pete Hegseth cited national security concerns to brand Anthropic a supply chain risk. Anthropic CEO Dario Amodei refused to acquiesce amid concerns the company's technology could be applied to mass surveillance or autonomous military drones. In her ruling, Lin stated that federal actions "were based on a desire to make a public example out of Anthropic for its 'arrogance' in criticizing the government, not based on any articulable basis to believe that Anthropic would actually sabotage its model." An Anthropic spokesperson welcomed the outcome in a statement: "We remain focused on working productively with the government to harness AI for our national security so all Americans benefit from this technology." The White House did not immediately issue a response to requests for comment. Writing in the 59-page decision, Lin, an appointee of former President Joe Biden, asserted that constitutional provisions and federal statutes do not allow authorities to "impose sweeping penalties based principally on Anthropic's critique of the Administration's views." Anthropic launched its lawsuit against the Pentagon in March in response to the supply chain classification, challenging what it described as an "unlawful campaign of retaliation" caused by its refusal to grant unrestricted military access to its tools. The court fight highlighted a rare public dispute regarding the role of AI in defense and surveillance operations. Notably, primary industry competitor OpenAI, creator of ChatGPT, reached its own agreement to collaborate with defense officials hours after the government penalized Anthropic. Both Anthropic and OpenAI are currently preparing for anticipated initial public offerings. Anthropic is also pursuing a secondary, narrower lawsuit currently pending in the federal appeals court in Washington, D.C., targeting a separate regulatory mechanism invoked by the Pentagon to designate the firm a risk. Earlier in the case, Lin granted a preliminary injunction prohibiting the Pentagon from applying the supply chain designation and halting the enforcement of a social media directive from Trump instructing federal agencies to cease using Anthropic products and its Claude chatbot. During a hearing on July 30, Lin described the government's stance as "really troubling" and remarked that it appeared "at odds to me with the First Amendment." She added that the legal standing had "gotten worse for the government" over time. Department of Justice attorneys argued at that hearing that the complex architecture of AI models is "so staggeringly enormous and opaque" that defense officials cannot evaluate it through standard methods used for physical hardware. Anthropic attorney Michael Mongan asserted during the proceeding that official actions "profoundly harm Anthropic" and "threaten more broadly to chill speech and debate on a very important issue."
US judge blocks Pentagon's Anthropic blacklisting
Add Yahoo as a preferred source to see more of our stories on Google. August 27(Reuters) - A U.S. judge on Thursday blocked the Pentagon's blacklisting of Anthropic, the latest turn in the Claude maker's high-stakes fight with the military over AI safety on the battlefield. Anthropic's lawsuit in California federal court alleges that Defense Secretary Pete Hegseth overstepped his authority when he designated Anthropic a national security supply-chain risk, a label the government can apply to companies that expose military systems to potential infiltration or sabotage by adversaries. Hegseth's unprecedented move, which followed Anthropic's refusal to allow the military to use AI chatbot Claude for U.S. surveillance or autonomous weapons, blocked Anthropic from certain military contracts. Anthropic executives have said it could cost the company billions of dollars in lost business and reputational harm. Anthropic says that AI models are not reliable enough to be safely used in autonomous weapons and that it opposes domestic surveillance as a violation of rights, but the Pentagon says private companies should not be able to constrain military action. U.S. District Judge Rita Lin, an appointee of former Democratic President Joe Biden, made the ruling in a 59-page order where she found that the Pentagon's decision was "illegal and baseless." "The empty invocation of national security is not a blank check to punish and retaliate against government critics," she wrote. Anthropic's designation was the first time a U.S. company has been publicly designated a supply-chain risk under an obscure government-procurement statute aimed at protecting military systems from foreign sabotage. In its March 9 lawsuit, Anthropic alleged the government violated its right to free speech under the First Amendment of the Constitution by retaliating against its views on AI safety. The company said it was not given a chance to dispute the designation, in violation of its Fifth Amendment right to due process. The lawsuit says the decision was unlawful, unsupported by facts and inconsistent with the military's past praise of Claude. The Justice Department countered that Anthropic's refusal to lift the restrictions could cause uncertainty in the Pentagon over how it could use Claude and risk disabling military systems during operations, according to a court filing. The government said the designation stemmed from Anthropic's refusal to accept contractual terms, not its views on AI safety. Anthropic has a second lawsuit pending in Washington, D.C., over a separate Pentagon supply-chain risk designation that could lead to its exclusion from civilian government contracts. (Reporting by Jack Queen in New York; Editing by Noeleen Walder and Matthew Lewis)
US judge blocks Pentagon's Anthropic blacklisting
August 27(Reuters) - A U.S. judge on Thursday blocked the Pentagon's blacklisting of Anthropic, the latest turn in the Claude maker's high-stakes fight with the military over AI safety on the battlefield. Anthropic's lawsuit in California federal court alleges that Defense Secretary Pete Hegseth overstepped his authority when he designated Anthropic a national security supply-chain risk, a label the government can apply to companies that expose military systems to potential infiltration or sabotage by adversaries. Hegseth's unprecedented move, which followed Anthropic's refusal to allow the military to use AI chatbot Claude for U.S. surveillance or autonomous weapons, blocked Anthropic from certain military contracts. Anthropic executives have said it could cost the company billions of dollars in lost business and reputational harm. Anthropic says that AI models are not reliable enough to be safely used in autonomous weapons and that it opposes domestic surveillance as a violation of rights, but the Pentagon says private companies should not be able to constrain military action. U.S. District Judge Rita Lin, an appointee of former Democratic President Joe Biden, made the ruling in a 59-page order where she found that the Pentagon's decision was "illegal and baseless." "The empty invocation of national security is not a blank check to punish and retaliate against government critics," she wrote. Anthropic's designation was the first time a U.S. company has been publicly designated a supply-chain risk under an obscure government-procurement statute aimed at protecting military systems from foreign sabotage. In its March 9 lawsuit, Anthropic alleged the government violated its right to free speech under the First Amendment of the Constitution by retaliating against its views on AI safety. The company said it was not given a chance to dispute the designation, in violation of its Fifth Amendment right to due process. The lawsuit says the decision was unlawful, unsupported by facts and inconsistent with the military's past praise of Claude. The Justice Department countered that Anthropic's refusal to lift the restrictions could cause uncertainty in the Pentagon over how it could use Claude and risk disabling military systems during operations, according to a court filing. The government said the designation stemmed from Anthropic's refusal to accept contractual terms, not its views on AI safety. Anthropic has a second lawsuit pending in Washington, D.C., over a separate Pentagon supply-chain risk designation that could lead to its exclusion from civilian government contracts. (Reporting by Jack Queen in New York; Editing by Noeleen Walder and Matthew Lewis)
