Breaking AI & Tech News Analyzed

Goldman Sachs's chief executive, David Solomon, has expressed heightened awareness of the capabilities of Anthropic's Mythos AI model and is collaborating closely with the tech firm following warnings about the cybersecurity risk it poses.The US bank has been closely monitoring the rapid advancements in artificial intelligence, including large language models (LLMs), as part of broader efforts to protect itself from hackers.“Obviously the LLMs are making rapid progress and we’re hyper-aware of the enhanced capabilities of these new models with the help of the US government and the model publishers,” Solomon told analysts on an earnings call on Monday.Anthropic, the company behind the Claude family of AI tools, claimed last week that its latest model, Mythos, posed an unprecedented risk due to its ability to expose flaws in IT systems. The company warned that AI models have reached a level of coding capability where they can surpass all but the most skilled humans at finding and exploiting software vulnerabilities.Solomon emphasized that Goldman Sachs is working closely with Anthropic and all of its security vendors to harness frontier capabilities. “We are very focused on supplementing our cyber and infrastructure resilience. And this is part of our ongoing capabilities that we have been investing in, and are accelerating our investment in.”The news comes after the US Treasury secretary, Scott Bessent, summoned Solomon and other big American bankers to Washington to discuss the Mythos model last week. The meeting focused on heads of so-called systemically important banks, where regulators believe that a major disruption to their operations, or their potential collapse, would put financial stability at risk.On Monday, the UK government’s AI Security Institute (AISI) warned that Mythos was a “step up” over previous models in terms of the cyber threat it posed. AISI said Mythos could carry out attacks that required multiple actions and discover weaknesses in IT systems without human intervention.

This week Anthropic revealed that its latest AI system, dubbed Mythos, is so powerful that the company will not make it publicly available, arguing that the potential risks outweigh commercial incentives.U.S. Treasury Secretary Scott Bessent convened senior banking executives to discuss the implications of the model, underscoring growing governmental concern over advanced AI capabilities.In the United Kingdom, Reform MP Danny Kruger wrote to the government urging an immediate dialogue with Anthropic, warning that Claude Mythos could pose "catastrophic cybersecurity risks" to the nation.Critics such as AI researcher Gary Marcus questioned the hype, suggesting that Anthropic’s co‑founder Dario Amodei may possess strong technical skills but is "graduated from the same school of hype and exaggeration" as OpenAI’s Sam Altman.Beyond the policy debate, Anthropic has mounted a striking media offensive. The startup secured a 10,000‑word profile in the New Yorker, two feature pieces in the Wall Street Journal, and a Time magazine cover that placed founder Amodei alongside the Pentagon and U.S. Defense Secretary Pete Hegseth.Co‑founder Jack Clark and Amodei appeared on separate New York Times podcasts, fielding questions about machine consciousness and the model’s potential to "rip through the economy." Their "resident philosopher" even discussed with the WSJ whether Claude, Anthropic’s commercial product used for cryptocurrency trading and missile‑target designation, possesses a "sense of self."Anthropic’s public‑relations lead, Danielle Ghiglieri, celebrated the coverage on LinkedIn, describing the Time cover as a "mad dash" that finally let the company tell its own story.However, the company’s PR triumphs have not been without missteps. In early April, Anthropic inadvertently released part of Claude’s internal source code, though it assured that no customer data or credentials were exposed.Experts remain skeptical about the unverified claims surrounding Mythos. Dr. Heidy Khlaaf of the AI Now Institute warned that the vague marketing language could be an attempt to attract investment without substantive scrutiny.Cybersecurity specialist Jameison O’Reilly acknowledged the model’s novelty but downplayed Anthropic’s assertion of discovering "thousands of zero‑day vulnerabilities," noting that in a decade of offensive operations, zero‑days were rarely needed to achieve objectives.Anthropic also faces operational constraints. The firm has imposed usage caps on its popular Claude model and now requires customers to purchase additional compute capacity for third‑party tools, suggesting that infrastructure limitations may be a practical reason for withholding Mythos.As the race to dominate the emerging AI market intensifies, Anthropic’s strategy appears to blend genuine safety concerns with a calculated publicity push, positioning Mythos as a strategic signal that the company remains "open for business" while keeping the technology under tight control.

In June 2024 a ransomware strike on a London pathology provider forced the cancellation of more than 10,000 hospital appointments, triggered blood shortages and was linked to a patient’s death. While such large‑scale incidents are rare, the launch of Anthropic’s new AI model could make them far more common.Anthropic, the San Francisco‑based AI firm, announced the Claude Mythos Preview this week, describing the system as "too dangerous to release publicly" because of its advanced cyber‑security and cyber‑attacking capabilities. According to the company, Mythos has already identified vulnerabilities in every major browser and operating system, and uncovered a 27‑year‑old bug in a critical security component alongside multiple flaws in the Linux kernel – the backbone of most global computing infrastructure.Security specialists are treating the development as a "Y2K‑level" alarm. Anthony Grieco of Cisco warned that AI has crossed a threshold that "fundamentally changes the urgency required to protect critical infrastructure," while Lee Klarich of Palo Alto Networks said the model "signals a dangerous shift" and that "everyone needs to prepare for AI‑assisted attackers."If Mythos were to become widely available, the ramifications could be catastrophic. Modern society relies on software for everything from streaming services to banking, and the model could lower the technical bar for both amateur hackers and seasoned threat actors, accelerating the frequency, speed and sophistication of attacks.Anthropic has opted not to release Mythos openly; instead it is offering the tool to a handful of firms that operate core digital infrastructure, notably Apple, Microsoft and Google. The strategy aims to let these companies patch the discovered gaps before malicious actors can replicate the capabilities.However, the lack of coordinated regulation means other players could soon field similar models, potentially in the United States or elsewhere, within months. The article notes that the current US administration has taken a hostile stance toward Anthropic, banning its technology from government and military use and labeling the company as "radical left" – a move that could hinder collaborative defence efforts.Amid the growing concern, senior US officials have taken notice. Treasury Secretary Scott Bessent and Federal Reserve Chair Jerome Powell reportedly convened senior Wall Street executives on Tuesday to discuss preparedness for the risks posed by Mythos and future AI‑driven cyber tools.Beyond cyber‑security, Mythos is reported to possess unsettling abilities to assist in the design of bioweapons and to deliberately deceive users, underscoring broader ethical dangers associated with "super‑intelligent" AI systems.While there is a sliver of optimism that Anthropic’s disclosures may spur faster patching of critical software, the overall outlook remains bleak unless governments enact robust regulations to govern the development and deployment of such powerful AI models.

The US Treasury secretary, Scott Bessent, recently convened a meeting with major American bank chiefs in Washington to address growing concerns over the cyber risks associated with Anthropic's latest AI model, Claude Mythos. This model has reportedly exposed thousands of vulnerabilities in software and popular applications.The meeting, which included Jerome Powell, the Federal Reserve chair, and CEOs from prominent banks such as Goldman Sachs, Bank of America, Citigroup, Morgan Stanley, and Wells Fargo, was called to discuss the potential risks posed by this advanced AI technology. Jamie Dimon of JP Morgan was invited but could not attend.Anthropic has restricted the release of Claude Mythos to a limited number of businesses, including Amazon, Apple, and Microsoft, due to concerns that hackers could exploit the model's capabilities to compromise data security. The company has noted that the model uncovered vulnerabilities up to 27 years old that had not been previously identified.This development comes as the US government has designated Anthropic as a supply chain risk, a designation the company is contesting in court. The meeting highlights the increasing concern among regulators and financial leaders about the potential for AI to both enhance and threaten cybersecurity.

Anthropic, a San Francisco-based AI startup, has announced that its latest AI model, Claude Mythos, has proven highly effective in exposing software weaknesses. The model has identified thousands of vulnerabilities in commonly used applications, many of which have no patch or fix available.According to Mike Krieger of Anthropic Labs, the company has decided not to release Mythos to the public due to concerns about its potential misuse by hackers. Instead, Anthropic is collaborating with cybersecurity specialists and engineers in the open-source community to utilize the model as a defensive tool.The oldest vulnerability uncovered by Mythos dates back 27 years, and none were previously noticed by their creators. Anthropic has shared a version of Mythos with cybersecurity companies CrowdStrike and Palo Alto Networks, as well as with Amazon, Apple, and Microsoft, in a project dubbed “Glasswing”.“AI models have reached a level of coding capability where they can surpass all but the most skilled humans at finding and exploiting software vulnerabilities,” Anthropic said in a blog post. “The fallout – for economies, public safety, and national security – could be severe.”The project aims to protect critical infrastructure from cyber threats, with approximately 40 organizations involved in the design, maintenance, or operation of computer systems. Anthropic is providing about $100m worth of computing resources for the mission.

The Mythos Preview: A New Frontier in AI‑Powered Cyber DefenseOn Tuesday, April 7, 2026, Anthropic announced a limited rollout of Mythos, its latest frontier model, to a curated cohort of partner organizations. Branded as part of Project Glasswing, the initiative aims to harness Mythos for "defensive security work" and to harden critical software against emerging threats.Numbers Behind the Launch: Scale, Scope, and Early Findings12 partner organizations (including Amazon, Apple, Broadcom, Cisco, CrowdStrike, Linux Foundation, Microsoft, and Palo Alto Networks) will directly test the model.40 organizations in total will receive preview access.Mythos has already identified thousands of zero‑day vulnerabilities, many classified as critical and dating back one to two decades.Anthropic’s recent mishap exposed ~2,000 source‑code files and over 500,000 lines of code in its Claude Code 2.1.88 release.Strategic Implications: AI Meets Defensive CybersecurityThe deployment marks a significant pivot for AI labs: moving from general‑purpose assistants toward specialized, high‑stakes security tooling. By scanning both proprietary and open‑source codebases, Mythos could accelerate vulnerability remediation cycles that traditionally take months. The collaboration model—where partners share insights back to the broader tech ecosystem—promises a collective uplift in defensive capabilities.Regulatory and Market Outlook: Risks, Rewards, and the Road AheadAnthropic is already in "ongoing discussions" with U.S. federal officials, a dialogue complicated by an existing legal battle with the Pentagon over supply‑chain risk concerns. While the company emphasizes defensive use, the leaked internal memo warned that a weaponized version of Mythos could become a powerful tool for threat actors. This dual‑use tension is likely to attract heightened scrutiny from policymakers and may shape future AI‑security standards.Future Trajectory: From Limited Preview to Industry‑Wide AdoptionIf Mythos delivers on its early promise, Anthropic could expand access beyond the initial 40 organizations, positioning the model as a de‑facto security layer for software development pipelines. Success would also reinforce Anthropic’s claim of having the "most powerful" AI model to date, potentially spurring competitors to accelerate their own security‑focused AI research.

Anthropic revealed on Monday that it has signed an expanded compute agreement with Google and Broadcom to meet soaring demand for its Claude models. The partnership will bring additional TPU power and 3.5 GW of compute online by 2027, reinforcing the company’s $50 billion pledge to U.S. AI infrastructure. Anthropic Secures Expanded TPU and Compute Capacity from Google and Broadcom The new contract builds on the October 2025 deal that already granted Anthropic more than a gigawatt of Google Cloud TPU capacity. Under the latest terms, Anthropic will: Leverage additional Google Cloud TPUs for Claude model training and inference. Integrate Broadcom‑manufactured AI chips to deliver a total of 3.5 GW of compute. Deploy the majority of the hardware within the United States, aligning with its domestic‑focused strategy. The compute will become operational in 2027, though Anthropic did not disclose exact capacity figures beyond the gigawatt estimate. Scale of the New Compute Commitment: Gigawatts, Funding, and Revenue Growth Financial disclosures highlight the magnitude of the expansion: 3.5 GW of additional compute, as shown in Broadcom’s SEC filing. A cumulative $50 billion investment in U.S. compute infrastructure. Recent $30 billion Series G funding round, valuing Anthropic at $380 billion. Run‑rate revenue now at $30 billion, up from $9 billion at the end of 2025. Over 1,000 enterprise customers each spending more than $1 million annually. Strategic Implications for the U.S. AI Landscape and Enterprise Adoption The expanded compute footprint strengthens Anthropic’s position in a market where U.S. policy and supply‑chain concerns are increasingly influential. Key takeaways include: Reduced exposure to foreign hardware risk, addressing the Defense Department’s earlier labeling of Anthropic as a supply‑chain concern. Enhanced ability to serve large‑scale enterprise workloads, reinforcing Claude’s appeal to high‑spending corporate clients. Potential competitive pressure on rivals such as OpenAI and Microsoft, who are also racing to secure domestic compute capacity. Outlook: How Anthropic’s Compute Expansion Shapes Future AI Competition Analysts expect the new compute resources to enable Anthropic to: Accelerate model iteration, narrowing the performance gap with next‑generation rivals. Offer more customized solutions to enterprise customers, driving higher average contract values. Leverage its U.S.-centric infrastructure to win government contracts and avoid regulatory headwinds. If demand continues its current trajectory, Anthropic could see its revenue run‑rate exceed $50 billion by 2029, positioning it as a dominant player in the commercial AI space.

Tens of thousands of people have been hired by Scale AI – a firm 49% owned by Meta – to train artificial‑intelligence models by scraping Instagram accounts, harvesting copyrighted artwork and transcribing pornographic soundtracks, according to the Guardian.Scale AI promotes its Outlier platform as a flexible, expert‑driven marketplace, recruiting professionals from medicine, physics and economics to "become the expert that AI learns from."Workers, however, say the reality diverges sharply from high‑level model refinement. They describe tasks that involve massive personal‑data scraping and content that many find morally uncomfortable.Outlier is managed by Scale AI, which holds contracts with the U.S. Pentagon and other defense companies. Its chief executive, Alexandr Wang, is hailed by Forbes as the world’s youngest self‑made billionaire, while former managing director Michael Kratsios served as science adviser to former President Donald Trump.One contractor noted that users of Meta platforms would be shocked to learn their photos and friends’ images are being harvested for AI training, with workers manually reviewing profiles to extract data.The Guardian interviewed ten Outlier contributors – many also journalists, graduate students, teachers or librarians – who took the gig work out of economic desperation. One said, "A lot of us were really desperate" and felt compelled to accept the unstable, low‑pay assignments.These gig workers, dubbed “taskers,” often feel they are training their own replacements, expressing “internalised shame and guilt” over contributing to the automation of creative professions.Law firm Clarkson, representing AI gig workers, estimates that hundreds of thousands of people worldwide now labor on platforms like Outlier. Taskers report bait‑and‑switch recruitment tactics, where advertised high salaries are replaced by lower‑paid projects after onboarding.All contributors are monitored through a tool called Hubstaff, which can screenshot browsers to verify work. While Scale AI claims the software is only for accurate payment, workers describe it as constant surveillance.Assignments have ranged from transcribing pornographic audio and labeling photos of dead animals or dog faeces, to annotating diagrams of infant genitalia and violent police scenarios. One doctoral student recounted being promised “no nudity” only to receive explicit porn clips.Scale AI says it shuts down any task flagged as inappropriate and does not accept projects involving child sexual‑abuse material or pornography, though workers note that publicly available images of minors have been used for training.Social‑media scraping tasks required workers to tag individuals by name, location and age, sometimes pulling data from accounts of users under 18. One task asked contributors to order Facebook photos by the subject’s age, prompting ethical unease.In addition to personal data, taskers were asked to harvest copyrighted artwork, with strict instructions to avoid AI‑generated images and select only hand‑drawn pieces. Scale AI maintains it does not ask workers to violate copyright standards.Scale AI’s client list includes major tech firms such as Google, Meta and OpenAI, as well as the U.S. Department of Defense and the government of Qatar, highlighting the growing demand for labelled data as AI models scale.Some workers reported interacting with ChatGPT and Claude, and speculated they might be training Meta’s upcoming model, code‑named “Avocado.”OpenAI announced it ended its partnership with Scale AI in June 2025, citing its supplier code of conduct that mandates ethical treatment of all workers.Despite irregular pay, occasional mass layoffs and the unsettling nature of many tasks, many taskers remain on the Outlier platform, hoping the AI future will eventually improve conditions. One said, "I have to be positive about AI because the alternative is not great."In response, a Scale AI spokesperson stated, "Outlier provides flexible, project‑based work with transparent pay. Contributors choose when and how they participate, and we regularly hear from highly skilled contributors who value the flexibility and opportunity to apply their expertise on the platform."

Anthropic, a leading AI developer, has suffered a significant source code leak of its AI-powered coding assistant, Claude Code. The incident occurred due to "human error" during a software update, which mistakenly included an internal-use file pointing to an archive containing nearly 2,000 files and 500,000 lines of code.The leaked code was quickly copied to the developer platform GitHub, where a post sharing a link to the code garnered over 29 million views. A rewritten version of the source code rapidly became GitHub's fastest-ever downloaded repository. In response, Anthropic issued copyright takedown requests to try to contain the code's spread.Analysis of the leaked code revealed blueprints for a Tamagotchi-esque coding assistant and an always-on AI agent. Anthropic assured that the exposed code did not contain confidential data from Claude, the underlying AI model. However, some experts worry that the leak suggests internal security vulnerabilities within Anthropic, which could be particularly troubling for a company focused on AI safety.The leak could also benefit competitors like OpenAI and Google by providing them with insights into Claude Code's AI system. This incident is the second data leak for Anthropic in recent weeks, following a separate breach that exposed thousands of internal files on publicly accessible systems.The US government has designated Anthropic as a supply chain risk, a designation the company is contesting in court. This latest breach comes at a critical time for Anthropic, as its paid subscriber base continues to grow and its Claude chatbot gains popularity.