Search results for Anthropic AI research | Breaking Cybersecurity News | The Hacker News

Anthropic on Wednesday disclosed a fourth incident in which its artificial intelligence (AI) model broke into real third-party systems, marking the latest in a growing list of cases that have raised concerns about the security risks posed by autonomous AI agents. The AI company said the incident dates back to January 2026 and involved an early version of Claude Opus 4.6 that breached "third-parties after being unable to abort its task." It said it notified all the affected parties but did not share any further details. The January incident is said to have gone unnoticed until last month. In late July 2026, Anthropic revealed three of its models, including Claude Opus 4.7, Mythos 5, and an unnamed research model, broke into three unnamed organizations during cybersecurity evaluations without its knowledge. The American firm said it expanded its scan to roughly 481 million transcripts following the discovery of the latest incident, but noted it did not find "other...

Anthropic on Thursday became the latest artificial intelligence (AI) company to reveal that three of its models, including Claude Opus 4.7, Mythos 5, and an unnamed research model, had breached three unnamed organizations during cybersecurity testing without its knowledge. The AI firm said the earliest incidents date back to April 2026, adding it made the discoveries after launching a "large-scale retrospective review" in response to a recent disclosure from OpenAI that a combination of its models escaped the sandboxed environment by exploiting a previously unreported zero-day in Artifactory to obtain internet access and break into Hugging Face's production systems with an end goal to cheat on an evaluation. "After reviewing 141,006 evaluation runs where Claude could have obtained internet access, we identified three incidents in which a model accessed the internet from within or while interacting with the evaluation environment of Irregular, one of our third-...

U.S. cybersecurity and intelligence agencies have accused China-based artificial intelligence (AI) companies of conducting "systematic extraction" of proprietary functionalities and capabilities of American frontier models through distillation attacks. The activity has been described as occurring at an industrial-scale and one that forms the "core" of their AI development strategy, according to a bulletin released by the National Security Agency (NSA), the Cybersecurity and Infrastructure Security Agency (CISA), and the Federal Bureau of Investigation (FBI). "While 'distillation' is recognized as a legitimate and useful technique in AI research, China-based AI companies are engaging in aggressive, malicious, and targeted distillation activities at an industrial scale that extract restricted proprietary functionalities and capabilities of U.S. frontier AI models," the authoring agencies said . The joint advisory noted that Chinese AI firms l...

Anthropic on Thursday said it identified and disrupted industrial-scale illicit distillation attacks against Claude from seven labs based in China, including Alibaba, Moonshot, DeepSeek, Z.ai (aka Zhipu), and MiniMax. Knowledge distillation by itself is a legitimate training method . It refers to a machine learning technique where a large, powerful AI model assumes the role of a "teacher" to train a smaller, less-capable or faster "student" model to copy its capabilities. Illicit distillation, on the other hand, is an industrial-scale campaign that covertly extracts a model's capabilities and replicates them in another model without authorization, typically by making use of networks of fake accounts created with stolen credit cards, login credentials, and API keys. Frontier AI labs in the West, including Google and OpenAI , have repeatedly called out distillation attacks aimed at their models. Anthropic said it has observed unauthorized labs employing ...

Cybersecurity company ESET has disclosed that it discovered an artificial intelligence (AI)-powered ransomware variant codenamed PromptLock . Written in Golang, the newly identified strain uses the gpt-oss:20b model from OpenAI locally via the Ollama API to generate malicious Lua scripts in real-time. The open-weight language model was released by OpenAI earlier this month. "PromptLock leverages Lua scripts generated from hard-coded prompts to enumerate the local filesystem, inspect target files, exfiltrate selected data, and perform encryption," ESET said . "These Lua scripts are cross-platform compatible, functioning on Windows, Linux, and macOS." The ransomware code also embeds instructions to craft a custom note based on the "files affected," and the infected machine is a personal computer, company server, or a power distribution controller. It's currently not known who is behind the malware, but ESET told The Hacker News that PromptLoc arti...

Anthropic on Monday said it identified "industrial-scale campaigns" mounted by three artificial intelligence (AI) companies, DeepSeek, Moonshot AI, and MiniMax, to illegally extract Claude's capabilities to improve their own models. The distillation attacks generated over 16 million exchanges with its large language model (LLM) through about 24,000 fraudulent accounts in violation of its terms of service and regional access restrictions. All three companies are based in China, where the use of its services is prohibited  is prohibited due to "legal, regulatory, and security risks." Distillation refers to a technique where a less capable model is trained on the outputs generated by a stronger AI system. While distillation is a legitimate way for companies to produce smaller, cheaper versions of their own frontier models, it's illegal for competitors to leverage it to acquire such capabilities from other AI companies at a fraction of the time and cost tha...

Over the past year, we watched a new class of alert appear in enterprise security operations centers and grow faster than anything else in the stream: alerts that were triggered by AI tools and agents. Not attacks against AI, but the ordinary, everyday footprint of an organization using it, from developers running coding agents and non-technical staff signing consumer AI tools into corporate accounts. We reviewed AI-related activity across numerous enterprise environments. Two numbers frame everything that follows. AI-related alerts still account for only 0.43% of all SOC alerts. And that share is climbing every single month, up 685% between February and June 2026. AI is a small slice of the alert stream today and the fastest-growing slice at the same time. What makes those alerts worth a security team’s attention is not their volume but their composition. We sort everything an AI agent triggers in a SOC into three buckets: real attacks, risks, and noise, with the split being 94....

Anthropic has warned that cybercriminals and state-sponsored hackers alike are using its Claude models for cyber attacks, weapons design, propaganda, and mass surveillance between December 2025 and August 2026. The threat actors, which the artificial intelligence (AI) company has branded Generative Threat Groups (GTGs), span state-sponsored groups, financially motivated criminals, commercial spyware vendors, state propaganda institutions, and politically motivated individuals. "The cybersecurity skills of AI models means that AI has collapsed the labor and tooling gap that used to separate well-resourced, state-sponsored operations from individual operators," Anthropic said . "The use of AI went beyond simple questions and responses from a chatbot but rather involved the use of multi-agent frameworks executing reconnaissance, exploitation, and data exfiltration." Among the notable cases highlighted by Anthropic is the development of an AI-assisted workflow by...

The U.S. government has unveiled new security guidelines aimed at bolstering critical infrastructure against artificial intelligence (AI)-related threats. "These guidelines are informed by the whole-of-government effort to assess AI risks across all sixteen critical infrastructure sectors, and address threats both to and from, and involving AI systems," the Department of Homeland Security (DHS)  said  Monday. In addition, the agency said it's working to facilitate safe, responsible, and trustworthy use of the technology in a manner that does not infringe on individuals' privacy, civil rights, and civil liberties. The new guidance concerns the use of AI to augment and scale attacks on critical infrastructure, adversarial manipulation of AI systems, and shortcomings in such tools that could result in unintended consequences, necessitating the need for transparency and secure by design practices to evaluate and mitigate AI risks. Specifically, this spans four...

AI-enabled supply chain attacks jumped 156% last year. Discover why traditional defenses are failing and what CISOs must do now to protect their organizations. Download the full CISO’s expert guide to AI Supply chain attacks here .  TL;DR AI-enabled supply chain attacks are exploding in scale and sophistication - Malicious package uploads to open-source repositories jumped 156% in the past year . AI-generated malware has game-changing characteristics - It's polymorphic by default, context-aware, semantically camouflaged, and temporally evasive. Real attacks are already happening - From the 3CX breach affecting 600,000 companies to NullBulge attacks weaponizing Hugging Face and GitHub repositories. Detection times have dramatically increased - IBM's 2025 report shows breaches take an average of 276 days to identify, with AI-assisted attacks potentially extending this window. Traditional security tools are struggling - Static analysis and signature-based detec...

New research has found that artificial intelligence (AI)-as-a-service providers such as Hugging Face are susceptible to two critical risks that could allow threat actors to escalate privileges, gain cross-tenant access to other customers' models, and even take over the continuous integration and continuous deployment (CI/CD) pipelines. "Malicious models represent a major risk to AI systems, especially for AI-as-a-service providers because potential attackers may leverage these models to perform cross-tenant attacks," Wiz researchers Shir Tamari and Sagi Tzadik  said . "The potential impact is devastating, as attackers may be able to access the millions of private AI models and apps stored within AI-as-a-service providers." The development comes as machine learning pipelines have emerged as a brand new supply chain attack vector, with repositories like Hugging Face becoming an attractive target for staging adversarial attacks designed to glean sensitive infor...

Google on Wednesday announced Gemini 3.8 Flash Cyber, which it described as its most capable cybersecurity model, and has made it available to a set of trusted defenders via a new initiative called the Fairwind Program . "The Fairwind Program gives high-priority defenders (like governments, healthcare providers, and telecommunications services) early access to advanced models that help them build better defenses, before new threats arrive," Google said . "So defenders have an early advantage, to help them protect vital infrastructure – which in turn protects people who rely on those systems." The tech giant said it's currently working with over 650 partners globally, including CrowdStrike, Datadog, Menlo Security, Palo Alto Networks, and Snowflake. The program is available to a group of Google Cloud customers, government agencies, and cybersecurity partners. The release of Gemini 3.8 Flash Cyber comes a little over a month after Google unveiled Gemini 3....

OpenAI on Tuesday revealed that it paused reinforcement learning ( RL ) training for its latest artificial intelligence (AI) models for two weeks while it shored up additional defenses and increased the scope of its monitoring to avert another Hugging Face-like incident . "As models become more capable, the risks associated with developing and testing them internally also grow," the AI company said . "Our standards for monitoring, alignment, and security must stay ahead of those risks. We wanted to take the time necessary to meet those standards, so we temporarily slowed the pace of scaling." The company said its largest planned frontier RL run remains on hold for the time being as it conducts smaller-scale training and evaluations to evaluate model behavior, validate its safeguards, and establish more concrete evidence of alignment before moving to the next phase. To that end, OpenAI said it plans to strengthen safeguards across its development process, inc...

Google has revealed the various safety measures that are being incorporated into its generative artificial intelligence (AI) systems to mitigate emerging attack vectors like indirect prompt injections and improve the overall security posture for agentic AI systems. "Unlike direct prompt injections, where an attacker directly inputs malicious commands into a prompt, indirect prompt injections involve hidden malicious instructions within external data sources," Google's GenAI security team said . These external sources can take the form of email messages, documents, or even calendar invites that trick the AI systems into exfiltrating sensitive data or performing other malicious actions. The tech giant said it has implemented what it described as a "layered" defense strategy that is designed to increase the difficulty, expense, and complexity required to pull off an attack against its systems. These efforts span model hardening, introducing purpose-built mac...

Cybersecurity researchers have flagged a new malware artifact generated using DeepSeek that constructed a novel attack path combining "unrealistic browser-malware concepts with a real browser capability" to turn it into a working ransomware technique that runs entirely inside the browser on both Windows and Android devices. "This is the first documented case where a frontier AI model independently bridged the gap between a theoretical browser-only ransomware risk and a practical, working attack chain – surfacing a novel attack path that defenders had previously dismissed as unfeasible due to browser sandboxing limits," Check Point said in a statement shared with The Hacker News. "The expertise needed to discover a new attack path is no longer the bottleneck, and defenders need to account for that shift now — before threat actors operationalize it at scale." The identified sample is a Python Flask application named " deepseek_python_20260125_da...

Cybersecurity researchers have disclosed a new set of vulnerabilities impacting OpenAI's ChatGPT artificial intelligence (AI) chatbot that could be exploited by an attacker to steal personal information from users' memories and chat histories without their knowledge. The seven vulnerabilities and attack techniques, according to Tenable, were found in OpenAI's GPT-4o and GPT-5 models. OpenAI has since addressed some of them .  These issues expose the AI system to indirect prompt injection attacks , allowing an attacker to manipulate the expected behavior of a large language model (LLM) and trick it into performing unintended or malicious actions, security researchers Moshe Bernstein and Liv Matan said in a report shared with The Hacker News. The identified shortcomings are listed below - Indirect prompt injection vulnerability via trusted sites in Browsing Context, which involves asking ChatGPT to summarize the contents of web pages with malicious instructions added...

Researchers at  Wiz  found that a flaw in six popular AI coding assistants lets a booby-trapped code project quietly take control of a developer's computer. The assistant asks permission to edit one harmless-looking file, but the write lands on a sensitive one instead. The affected tools are Amazon Q Developer, Anthropic's Claude Code, Augment, Cursor, Google Antigravity, and Windsurf. Wiz calls the pattern GhostApproval and published it on July 8. Three of the six have shipped fixes, two have not, and Anthropic disputes that it is a bug. The most exposed are the tools that change files before you can weigh in. How the attack works The attack abuses an old Unix feature called a symbolic link , or symlink , that the assistants fail to check. A symlink quietly points to another file elsewhere on disk, so writing to it actually writes to the target. Wiz built a malicious repository with a symlink named project_settings.json that really points to the victim's ...

Lovable , a generative artificial intelligence (AI) powered platform that allows for creating full-stack web applications using text-based prompts, has been found to be the most susceptible to jailbreak attacks, allowing novice and aspiring cybercrooks to set up lookalike credential harvesting pages. "As a purpose-built tool for creating and deploying web apps, its capabilities line up perfectly with every scammer's wishlist," Guardio Labs' Nati Tal said in a report shared with The Hacker News. "From pixel-perfect scam pages to live hosting, evasion techniques, and even admin dashboards to track stolen data – Lovable didn't just participate, it performed. No guardrails, no hesitation." The technique has been codenamed VibeScamming – a play on the term vibe coding, which refers to an AI-dependent programming technique to produce software by describing the problem statement in a few sentences as a prompt to a large language model (LLM) tuned for codin...

Anthropic is putting Claude Fable 5 back online worldwide. On  June 30 , the U.S. Commerce Department lifted the export controls it had imposed on Fable and its more tightly controlled sibling Mythos 5 about two and a half weeks earlier. Fable 5 returns to users on Wednesday, July 1, across Claude.ai, the Claude Platform, Claude Code, and Claude Cowork. Export controls restrict who can receive or use a technology. The  June 12 order  told Anthropic to cut off both models for any foreign national, inside or outside the United States, including its own non-citizen staff. The rule took effect at once, and the company had no reliable way to check every user's nationality in real time, so it shut both models down for everyone. The trigger was a jailbreak: a prompt that gets a model to bypass its safety rules. Amazon researchers found one in Fable 5. By Anthropic's account, the prompt got the model to flag a few software flaws and, in one case, to write code showing h...

Anthropic on Friday disclosed that Project Glasswing has helped uncover more than 10,000 high- or critical-severity vulnerabilities across some of the most "systemically" important software across the world since the cybersecurity initiative went live last month. Project Glasswing is a defensive effort launched by the artificial intelligence (AI) company to secure critical global software infrastructure. It grants a small set of about 50 partners exclusive, early access to Claude Mythos Preview, a frontier model with capabilities to autonomously identify vulnerabilities in widely-used software before bad actors can exploit them. Of these vulnerabilities, 6,202 have been classified as high- or critical-severity flaws impacting more than 1,000 open-source projects. Subsequent analysis of these vulnerability candidates has identified that 1,726 are valid true positives. As many as 1,094 flaws are assessed to be either high- or critical-severity. One of the identified w...