Coverage for this event has ended.<\/p>\n
3 hours ago<\/p>\n
Pinned<\/p>\n
DeSantis warns rogue AI agents could wipe out Americans’ bank accounts<\/p>\n
Florida Gov. Ron DeSantis<\/a> warned Thursday that rogue artificial intelligence agents could drain Americans’ bank accounts and investment portfolios, raising concerns about whether victims could recover their losses.<\/p>\n “Imagine a rogue AI agent liquidating your bank account and investment portfolio,” DeSantis wrote on X.<\/p>\n “If rogue agents liquidate hundreds of billions across millions of accounts, the AI company would go bankrupt and you wouldn\u2019t get made whole.”<\/p>\n DeSantis suggested criminal liability for AI executives could help deter such incidents, but warned that punishment after the fact would be insufficient.<\/p>\n His comments followed an Axios report that major U.S. banks have detected unauthorized third-party AI agents accessing customer accounts, with some attempting actions apparently independent of users’ instructions.<\/p>\n Posted by Brittany Miller<\/p>\n 2 hours agoMusic publisher sues Anthropic, Suno over alleged AI copyright infringement<\/p>\n Music publisher Reservoir Media filed separate federal lawsuits against AI companies Anthropic and Suno, accusing them of using copyrighted songs by Madonna, Johnny Cash and Joni Mitchell to train artificial intelligence<\/a> models without permission, Billboard reported Thursday.<\/p>\n The lawsuits, filed Wednesday, allege the companies copied protected music and lyrics to develop AI systems capable of generating new content that competes with human artists.<\/p>\n Reservoir cited songs including Joni Mitchell’s “Big Yellow Taxi,” Johnny Cash’s “Ring of Fire” and Madonna’s “Papa Don’t Preach” among the copyrighted works allegedly used without authorization.<\/p>\n “We disagree with the allegations and intend to defend this case vigorously,” an Anthropic spokesperson told Fox News Digital.<\/p>\n The legal action adds to mounting copyright challenges facing AI developers over their use of protected material for model training.<\/p>\n A representative for Suno did not immediately respond to Fox News Digital’s request for comment.<\/p>\n Posted by Brittany Miller<\/p>\n 3 hours agoNvidia CEO Jensen Huang urges Europe to invest $50B in AI data centers<\/p>\n Nvidia CEO Jensen Huang is urging European nations to accelerate artificial intelligence infrastructure development, warning that the continent risks falling further behind the U.S. and China in the global AI race.<\/p>\n Huang, touring Europe to promote AI investment, said a single gigawatt-scale data center could cost $50 billion to build but generate tens of billions in annual revenue.<\/p>\n “Each gigawatt datacenter is $50 billion. The revenue generation you expect is tens of billion dollars per year,” Huang said, estimating potential annual revenue of $40 billion from such infrastructure.<\/p>\n During a FOX Business discussion<\/a> Thursday, panelists questioned whether European countries have the entrepreneurial culture needed to capitalize on the opportunity.<\/p>\n “They lack venture culture,” one panelist said, arguing that Europe’s challenge extends beyond access to investment capital.<\/p>\n Posted by Brittany Miller<\/p>\n 3 hours agoHackers actively exploiting critical security flaw in VeloCloud network software<\/p>\n Arista Networks warned Thursday that hackers are actively exploiting a critical security vulnerability in VeloCloud Orchestrator software that could allow attackers to seize administrative control of affected systems.<\/p>\n The flaw, tracked as CVE-2026-114625<\/a>, carries a severity score of 9.9 out of 10 and allows attackers with even the lowest level of authenticated access to gain elevated privileges and impersonate a superuser.<\/p>\n Successful attacks could expose sensitive data, compromise network operations and give hackers access to internal functions normally restricted from remote users.<\/p>\n Arista said the vulnerability affects certain on-premises VeloCloud Orchestrator installations, while hosted versions have already been patched.<\/p>\n The company identified two IP addresses associated with attacks and urged customers to upgrade immediately to fixed software releases, restrict administrative access and review system logs for signs of compromise.<\/p>\n Posted by Brittany Miller<\/p>\n 4 hours agoAmericans divided on AI: 57% view technology as bad for society despite personal benefits<\/p>\n A majority of Americans view artificial intelligence as a threat<\/a> to broader society, even as half believe the technology delivers personal benefits to their daily lives, according to a nationwide ABC News\/Washington Post\/Ipsos poll released Wednesday.<\/p>\n The poll revealed 57% of adults say AI is bad for society<\/a> overall, compared to 42% who see it as a societal good.<\/p>\n A majority of Republicans (57%) say AI is personally good for them while a similar majority of Democrats say it’s personally bad for them (55%), marking a contrast between the majority political parties before the midterms.<\/p>\n The survey reveals key partisan and generational divides over tech adoption. While 63% of Democrats and 58% of independents view AI’s societal impact negatively, Republicans lean more favorable, with 51% calling it beneficial overall. <\/p>\n Anti-progressive anxiety remains centered<\/a> on widespread job displacement, cybersecurity risks, and economic inequality, alongside broad opposition to the local expansion of energy-intensive AI data centers powering the underlying infrastructure.<\/p>\n 5 hours agoMicrosoft: Chinese state-backed group weaponizes agentic AI in ConnectWise supply chain<\/p>\n Microsoft Security revealed<\/a> that Chinese state-affiliated threat actor Storm-0940 used an open-source, agentic AI penetration testing framework called ARTEX to execute a massive supply chain campaign targeting downstream ConnectWise customers. <\/p>\n The hack highlights the growing threat of autonomous agentic AI in nation-state cyber warfare.<\/p>\n The autonomous attack<\/a> framework allowed the threat group to automate multi-step cyber operations at machine speed, compressing intrusion timelines from initial access to full cloud takeover into less than three hours.<\/p>\n “Storm-0940 used ARTEX to reach more than 2,250 tenants at 2,900 sign-in attempts per minute, a pace that leaves defenders with little time to detect and respond,” according to Microsoft Threat Intelligence.<\/p>\n The campaign originated from a breach of ConnectWise’s Cloud Backup application environment, where Storm-0940 obtained privileged OAuth application credentials. <\/p>\n The attackers launched automated reconnaissance, password spraying, and OAuth credential modification routines across thousands of targeted Microsoft 365 customer tenants. <\/p>\n By leveraging pre-established trust relationships between ConnectWise and its client networks, the AI agents modified application credentials to assign themselves privileged Global Administrator roles and establish persistent cloud backdoors.<\/p>\n 5 hours agoSecurity researchers uncover 90 malicious AI skills using hidden-text evasions<\/p>\n Security researchers at VirusTotal<\/a> have identified 90 malicious AI agent skills designed to bypass human code review using stealthy text-formatting tricks. <\/p>\n While human reviewers inspecting the markdown files see standard documentation or harmless checklists, the underlying large language model (LLM)<\/a> ingests hidden prompt-injection instructions directly from the raw file.<\/p>\n 5 hours agoRhode Island sues Meta over AI-powered scam ads<\/p>\n Rhode Island has filed a lawsuit against Meta<\/a>, accusing the tech giant of using its automated AI ad creation tools to assist scammers in generating, optimizing, and deploying fraudulent marketing campaigns across Facebook, Instagram, and WhatsApp.<\/p>\n “Meta<\/a> knowingly helped create and optimize these ads to ensure that they were as effective as possible and then lied about them, misrepresenting to the public its commitment to prevent fraud on its platforms,” according to the Rhode Island lawsuit.<\/p>\n The complaint marks the first scams-related suit brought against Meta by a U.S. state. <\/p>\n Beyond exposing users to fraud, the state alleges that Meta’s automated systems<\/a> charged higher rates to advertisers suspected of fraud rather than blocking them, effectively profiting from high-risk scam campaigns that utilized AI to target vulnerable demographics.<\/p>\n 5 hours agoMeta’s credit risk hits record high as heavy AI spending spook debt markets<\/p>\n The cost of insuring Meta Platforms’ debt against default touched an all-time high Thursday as credit markets grow increasingly uneasy over Big Tech\u2019s debt-funded artificial intelligence expansion<\/a>. <\/p>\n Five-year credit default swaps (CDS) tied to Meta widened 7.6 basis points to a record 114.79 bps, marking its largest single-day surge since at least 2025, while Meta shares tumbled over 6% to six-week lows.<\/p>\n “The market has been questioning whether all the spending on AI is having the type of returns that the market originally expected,” Steve Sosnick, chief market analyst at Interactive Brokers, said. “In Meta’s case<\/a>, investors are interpreting the new non-voting shares as a backdoor for issuing more stock and diluting holdings eventually, so I think that’s adding to this aggressive selloff.”<\/p>\n The credit repricing follows Meta tightening its annual capital expenditure forecast upward to between $135 billion and $145 billion. <\/p>\n Heavy compute infrastructure investments resulted in a third-quarter free cash flow deficit of $2.12 billion \u2013 marking Meta’s first negative cash flow quarter since 2012. <\/p>\n The broader market shockwave extended across the tech sector, with the Philadelphia Semiconductor Index<\/a> shedding roughly $688 billion in market value while credit default swaps for peers like Oracle and Amazon held near historic peaks.<\/p>\n Reuters contributed to this report.<\/p>\n 5 hours agoOCC chief says banks are ‘all over this’ amid unauthorized AI agent access reports<\/p>\n Comptroller of the Currency Jonathan Gould analyzes the emerging artificial intelligence risks in the banking sector, including unauthorized AI agents accessing<\/a> customer accounts, in an appearance Thursday on FOX Business’ “Varney & Co.”<\/p>\n Gould also discussesed the regulatory frameworks for stablecoins finalized by the OCC.<\/p>\n 5 hours agoOpenAI discovers major security flaw that could impact millions<\/p>\n FOX Business national corespondent Darren Botelho joined “Fox & Friends” co-hosts Brian Kilmeade, Ainsley Earhardt and Lawrence Jones to analyze a massive internet security vulnerability discovered by OpenAI. <\/p>\n The critical PHP bug<\/a> poses a severe threat that could potentially take down 70% of the internet if left unpatched.<\/p>\n 6 hours agoNvidia launches Minerva platform for end-to-end AI factory observability<\/p>\n Nvidia<\/a> has unveiled Minerva, a specialized AI Ops platform designed to provide end-to-end operational visibility into massive AI data centers and supercomputing clusters. <\/p>\n By bridging the gap between hardware infrastructure and machine learning workloads, the software allows operators to identify performance bottlenecks and predict component failures before they disrupt large-scale AI jobs.<\/p>\n The platform<\/a> creates a dynamic knowledge graph that maps relationships across GPUs, networking switches, SuperNICs, and cables alongside active training workloads. <\/p>\n Analyzing billions of data points per hour<\/a>, Minerva groups individual hardware alerts into single root-cause incidents \u2013 allowing engineers to trace training slowdowns to specific faulty components across massive GPU environments while optimizing overall cluster uptime, token output, and compute yield.<\/p>\n 6 hours agoSenate defense leaders push for govt-controlled satellite constellations, automated drone war tech<\/p>\n Senate Armed Services Committee member Kevin Cramer,<\/a> R-N.D., highlighted a push to transition from private, commercial tech dependence toward sovereign, government-managed AI networks and automated manufacturing.<\/p>\n After a classified briefing with the Secretary of the Air Force regarding Ukraine’s access to SpaceX’s Starshield and Starlink satellite systems<\/a>, Cramer emphasized that while private sector innovation is crucial, operational control over low-Earth orbit (LEO) networks must sit firmly with the U.S. military.<\/p>\n “The operation of a low Earth satellite constellation for ISR for connectivity is really, really important that it be exclusively the authority of the United States government,” Cramer told reporters Thursday.<\/p>\n “What it demonstrates to me,” he added, “is that we cannot be so dependent on others, and fighting wars is one of those things.”<\/p>\n To address these vulnerabilities, Cramer highlighted the rapid deployment of the military\u2019s dedicated Space Data Network (SDN), noting that hundreds of secure satellites are currently heading into orbit to establish independent intelligence, surveillance, and reconnaissance (ISR) capabilities.<\/p>\n Cramer also spoke about automated manufacturing capabilities observed on the ground during a recent trip to Eastern Europe as a blueprint for Western defense.<\/p>\n “When I was in Ukraine last week, I saw a factory where they produce 5,000 weaponized drones<\/a> a day,” he said. “We need to figure out how we can do something like that, or partner with them … without Chinese components.”<\/p>\n 6 hours agoWalmart unveils tools to let AI agents shop for customers<\/p>\n Walmart announced a suite of new tools designed to allow autonomous AI agents to search products and make purchases on behalf of shoppers<\/a>. <\/p>\n The move signals a major push into agentic e-commerce<\/a>, drawing a sharp contrast with competitor Amazon, which has actively restricted third-party AI shopping bots from accessing its platform.<\/p>\n “The momentum and buy-in from top management teams seems to be making this into a real thing,” Mizuho analyst David Bellinger said. <\/p>\n Under the new system, Walmart is introducing “Agent Profiles,” allowing customers to set spending limits, monitor purchases separately, or require explicit approval for each transaction. <\/p>\n Additionally, the retailer launched the “Walmart Agent Gateway,” enabling major AI platforms<\/a> like OpenAI\u2019s ChatGPT, Google\u2019s Gemini, and Meta\u2019s Muse to connect directly to Walmart and Sam\u2019s Club systems. <\/p>\n Available to all account holders without requiring a Walmart+ membership, the tools restrict AI agents from directly accessing sensitive payment card data while granting Walmart the authority to block any bots violating safety rules.<\/p>\n Reuters contributed to this report.<\/p>\n 6 hours agoSenate permitting reform targets AI data center power demands<\/p>\n Upcoming Senate permitting reform legislation aims to protect consumer electricity bills while expanding the energy grid to meet the massive power demands of AI computing, according to National Energy Dominance Council Executive Director Jarrod Agen told FOX Business<\/a> on Thursday.<\/p>\n “One of the things we absolutely want in there is ratepayer protection,” Agen said on “Big Money in the Morning<\/a>,” saying the Trump administration supports AI development, but hyperscalers building heavy data center infrastructure must pay their own way. <\/p>\n “We all know there is a demand for more energy in this country. But if the hyperscalers are going to come in and build data centers, rates cannot go up for the American people. <\/p>\n “So part of the permitting reform has to have language, which it does now to protect the ratepayer against these huge data centers being built. <\/p>\n “The president is on board and wants to get more, more AI development, more super intelligent development, but it cannot be at the expense of ratepayers. <\/p>\n “This legislation makes it that we can build the infrastructure we need for baseload power<\/a>. The Senate needs to vote on it when they come back after the election, and we can help build more energy infrastructure in this country.”<\/p>\n 8 hours agoAnthropic now stands alone on fall IPO calendar as AI market jitters sideline rivals<\/p>\n Anthropic is pressing forward with plans for a mid-November initial public offering<\/a> (IPO), standing as the sole major AI firm remaining on a fall tech calendar otherwise emptied by rising interest rates and market volatility. <\/p>\n The AI lab is expected to make its prospectus public as early as late October and begin marketing shares to prospective investors during the week of Nov. 9.<\/p>\n Anthropic’s listing will test whether its central role in the AI sector \u2013 where it is eyeing a valuation of up to $1.5 trillion \u2013 can overcome broader market headwinds. <\/p>\n Rising interest rates<\/a> and growing skepticism around massive data center expenditures have forced peers to pull or delay their listings. <\/p>\n CoreWeave has suffered a steep selloff in shares and bonds, while infrastructure providers including Switch, Nscale, SB Energy, Elice, AirTrunk REIT, and Firmus have all paused or postponed their public offerings.<\/p>\n Reuters contributed to this report.<\/p>\n 8 hours agoKansas Dem candidate ‘not sure’ what AI regulations he supports<\/p>\n Kansas Democratic Senate candidate Adam Hamilton<\/a> admitted during a debate that he is “not sure” which specific federal AI regulations he would back, even as he urged lawmakers to establish an oversight agency modeled after the Nuclear Regulatory Commission.<\/p>\n “The reason why I’m not sure is because I don’t think anybody in Congress really fully understands the dangers of AI<\/a>,” Hamilton, a pastor and founder of the Church of the Resurrection in Leawood, argued.<\/p>\n Leaving AI oversight to tech companies is a “bad idea,” he added, calling on Congress to enlist industry experts to prevent threats to future generations. <\/p>\n Incumbent opponent Sen. Roger Marshall, R-Kan.,<\/a> advocated for a “light touch” regulatory approach focused on national security, domestic privacy, and staying ahead of China. <\/p>\n Marshall pushed back against broad global regulatory frameworks, warning that adversarial nations like Russia and China would never abide<\/a> by international rules that Democrats and globalists push.<\/p>\n 8 hours agoUS Commerce Dept awards $389 million for power semiconductors and fusion materials<\/p>\n The U.S. Commerce Department<\/a> announced plans Thursday to award $389 million in research and development funding to four tech companies advancing power semiconductor devices and critical materials. <\/p>\n Under the agreement, the federal government will receive minority, non-controlling equity stakes in exchange for the capital injection.<\/p>\n Drawn from a $10 billion R&D fund established by Congress in 2022<\/a>, the awards target next-generation chip technology for fusion power, AI hardware, and raw material purification. <\/p>\n Helion will receive up to $175 million for fusion-generator semiconductors, while Commonwealth Fusion Systems is slated for up to $85 million. Additionally, Vertical Semiconductor will obtain up to $73 million to advance gallium nitride chips for AI data centers, and Sibelco will receive up to $56 million for ultra-high-purity quartz processing.<\/p>\n Reuters contributed to this report.<\/p>\n 9 hours agoSpotify co-CEO Soderstrom says listening is not shifting to third-party AI agents<\/p>\n Spotify<\/a> Co-CEO Gustav S\u00f6derstr\u00f6m pushed back against claims that third-party AI agents will disrupt music streaming interfaces, asserting that user listening habits remain firmly rooted inside native music platforms. <\/p>\n Despite the rapid rise of autonomous background assistants<\/a> across personal finance and productivity apps, S\u00f6derstr\u00f6m emphasized that consumers prefer direct control over their audio curation and playback environments.<\/p>\n The executive’s comments highlight Spotify’s strategy to integrate proprietary AI features \u2013 such as its AI DJ, Prompted Playlists, and internal agent integrations \u2013 directly within its existing ecosystem rather than outsourcing discovery to external bots. <\/p>\n Reuters contributed to this report.<\/p>\n 9 hours agoRedington Gulf pairs with AWS to accelerate AI adoption across Middle East, Turkey, and Africa<\/p>\n Redington Gulf, a subsidiary of technology distributor Redington Ltd, has signed a strategic collaboration agreement with Amazon Web Services<\/a> (AWS) to drive cloud and artificial intelligence adoption across the Middle East, Turkey, and Africa (META) region.<\/p>\n Under the multi-year partnership, Redington Gulf will leverage AWS infrastructure and AI capabilities to help enterprise clients and small-to-medium businesses deploy and scale generative AI tools, cloud migration strategies, and advanced analytics. <\/p>\n The deal positions both companies to capture accelerating demand for cloud-native AI infrastructure<\/a> across rapidly expanding digital markets in the META region.<\/p>\n Reuters contributed to this report.<\/p>\n 9 hours agoeBay to slash 1,000 jobs in second major layoff round of the year<\/p>\n EBay is eliminating roughly 1,000 positions<\/a> \u2013 about 9% of its workforce \u2013 in its second major wave of job cuts this year as it pivots resources toward high-growth categories. <\/p>\n The reductions were outlined in an internal memo from CEO Jamie Iannone on Wednesday and confirmed by an eBay<\/a> spokesperson.<\/p>\n “Our strategy is working,” Iannone said. “We must decide where to direct our investments and resources to capitalize on future opportunities.” <\/p>\n The San Jose-based e-commerce platform previously cut 800 roles<\/a> (about 6% of staff) in February. <\/p>\n The new round of layoffs comes as eBay doubles down on luxury goods, collectibles<\/a>, refurbished items, and secondhand fashion \u2013 bolstered by its $1.4 billion acquisition of Depop from Etsy in July \u2013 ahead of its third-quarter earnings report next week.<\/p>\n EBay deploys AI tools to help sellers expedite posting of goods on its platform.<\/p>\n This is a summary of FOX Business’\u00a0Bradford Betz report.<\/p>\n 9 hours agoUS banks detect unauthorized AI agents in customer accounts, report says<\/p>\n Major U.S. banks are detecting unauthorized<\/a> third-party AI agents accessing customer accounts, with some bots attempting actions independent of user commands, according to an Axios report Thursday.<\/p>\n Banks including Morgan Stanley, Bank of America, JPMorgan, Wells Fargo and others are racing to contain agent risks.<\/p>\n AT&T customers will decide whether AI agents can manage their accounts, CEO John Stankey said Wednesday<\/a>, responding to Meta\u2019s new Muse Spark AI model<\/p>\n .Any account transactions made by an AI agent would still need to pass the company\u2019s validation and authentication checks, Stankey said.<\/p>\n The discovery comes as autonomous AI assistants<\/a> gain rapid consumer adoption across personal finance and work management applications.<\/p>\n Reuters contributed to this report.<\/p>\n 9 hours agoMeta bets big on AI with Muse agent app, but heavy infrastructure costs dent stock<\/p>\n Meta<\/a> delivered strong third-quarter revenue and rapid early adoption for its new AI agent, Muse, yet its stock slipped roughly 3% in premarket trading Thursday as investors confronted mounting capital expenses. <\/p>\n “We want to establish this Muse network, and we think that it’s going to be more valuable for people at scale when there are a lot of people and businesses using Muse,” Meta CEO Mark Zuckerberg<\/a> said.<\/p>\n The Instagram parent tightened its annual capital expenditure forecast to between $135 billion and $145 billion, resulting in a third-quarter free cash flow deficit of $2.12 billion as infrastructure spending outpaced operating cash.<\/p>\n To support its AI investments while preserving Zuckerberg’s voting control, Meta announced plans to issue new non-voting Class C common stock in January 2027 through a 2-for-1 split structure. <\/p>\n The spending push comes alongside severe legal hurdles, including a $25 billion penalty ordered by a New Mexico judge in a data-privacy case linked to Cambridge Analytica<\/a>, which Meta plans to appeal. <\/p>\n Despite the immediate financial strain and legal scrutiny, 13 brokerages raised their price targets on the stock, betting that Meta’s aggressive compute expansion<\/a> will drive core ad performance and long-term user engagement.<\/p>\n Reuters contributed to this report.<\/p>\n 10 hours agoBarclays to slash 5,000 jobs in AI efficiency push<\/p>\n Barclays<\/a> has revealed plans to eliminate roughly 5,000 roles \u2013 about 5% of its global workforce \u2013 as part of an artificial intelligence efficiency drive aimed at streamlining back-office operations. <\/p>\n Around 60% of the overall job cuts will impact staff in support functions such as developers, CFO Anna Cross told reporters, in one of the clearest signs yet from a big bank of how AI is enabling firms to reduce headcount in such roles.<\/p>\n The job cuts<\/a> were revealed as Barclays reported third-quarter profit of $3.08 billion, up 12% from a year ago but undershooting analysts’ forecasts as its fixed income trading business underperformed.<\/p>\n “I am disappointed, and you can be sure this will get my attention,” CEO C.S. Venkatakrishnan said about the bank’s traditional strength of fixed income trading, blaming a volatile quarter and weak trading in rates and credit products in particular.<\/p>\n The bank said it had taken further “structural cost actions” in the third quarter, part of its longer-term plan to reduce expenses like staff costs through greater use of technology.<\/p>\n Cross said 1,000 roles have already gone, with the remainder to be cut by the end of the year.<\/p>\n “The cuts illustrate just how significantly AI and automation could reshape the banking industry over the next few years,” according to Will Howlett, financials analyst at fund management business Quilter Cheviot.<\/p>\n Reuters contributed to this report.<\/p>\n 10 hours agoAnthropic unveils Claude Duet as race for autonomous AI agents heats up<\/p>\n Anthropic<\/a> has officially launched Claude Duet, a standalone AI agent app designed to shift artificial intelligence from reactive prompt-and-response to autonomous task execution.<\/p>\n Built on the company’s Opus 5.5 model, Duet acts as a continuous digital co-pilot for professionals and small-business owners, surfacing urgent items, drafting daily briefings, and executing complex workflows without waiting for user commands.<\/p>\n Currently available via a waitlisted research preview on iOS, Android, and web, Duet enters an increasingly competitive landscape against OpenAI\u2019s “dots” agents and Meta\u2019s Muse AI<\/a>.<\/p>\n 11 hours agoAnthropic’s Amodei warns ‘danger’ of AI, but seeking to help US lead ‘race to the top’<\/p>\n Anthropic CEO Dario Amodei<\/a> is once again sounding the alarm on the rapid acceleration of artificial intelligence, stressing that without rigorous safeguards and careful deployment, AI systems could pose significant threats to society.<\/p>\n \u201cIf not properly trained and carefully deployed, AI can be a danger in its own right,\u201d he wrote in the forward of the 20-page 2026 Public Benefit Report<\/a> released by the company.<\/p>\n Anthropic explicitly acknowledges the double-edged sword of its own technology, as innovation pushes are bringing about concerns for cybersecurity, public safety and future regulatory action.<\/p>\n Amodei has consistently advocated for a more measured approach, recently briefing the UN Security Council on catastrophic risks<\/a>, including the potential for advanced models to automate cyberattacks or aid in the creation of biological weapons if left unchecked.<\/p>\n “The same intelligence that powers protein binder discoveries could be used by malicious actors to engineer biological weapons,” he wrote. “Without appropriate safeguards and release structures, cyber capabilities could be exploited to carry out large-scale attacks instead of serving to strengthen our infrastructure.”<\/p>\n