• Farmers Insurance Exchange and its subsidiaries recently disclosed a significant security incident that compromised personal information of approximately 1.1 million customers through an unauthorized access to a third-party vendor’s database.

    The breach, which occurred on May 29, 2025, represents one of the largest insurance industry data exposures of the year, affecting customer records containing names, addresses, dates of birth, driver’s license numbers, and partial Social Security numbers.

    The attack timeline reveals a sophisticated intrusion that went undetected for approximately 24 hours before the vendor’s monitoring systems identified suspicious activity.

    On May 30, 2025, the unnamed third-party vendor alerted Farmers to the unauthorized database access, triggering immediate containment measures and blocking the threat actor.

    The vendor’s existing monitoring infrastructure proved crucial in limiting the exposure window, though investigators later confirmed that data acquisition had already occurred during the initial breach period.

    Following the incident discovery, Farmers analysts worked alongside external cybersecurity experts to conduct a comprehensive forensic investigation spanning nearly two months.

    The investigation revealed that the unauthorized actor had successfully penetrated the vendor’s database defenses and exfiltrated sensitive customer information before detection systems could intervene.

    Farmers researchers noted that the attack specifically targeted customer databases containing insurance policy holder information, suggesting a deliberate focus on high-value personal data.

    Database Infiltration and Persistence Mechanisms

    The attack vector analysis indicates the threat actor employed advanced persistent techniques to maintain unauthorized database access.

    While specific technical details remain undisclosed for security reasons, the prolonged investigation period suggests complex data extraction methods were utilized.

    The attacker’s ability to access and acquire substantial customer data within a compressed timeframe points to sophisticated database querying capabilities and potential privilege escalation within the vendor’s systems.

    Security experts noted that the incident highlights critical vulnerabilities in third-party vendor management, particularly regarding database access controls and real-time monitoring systems.

    The breach underscores the importance of implementing robust vendor security frameworks and continuous monitoring protocols to detect unauthorized database activities before data exfiltration occurs.

    Boost your SOC and help your team protect your business with free top-notch threat intelligence: Request TI Lookup Premium Trial.

    The post Farmers Insurance Cyber Attack – 1.1 Million Customers Data Exposed in Salesforce Attack appeared first on Cyber Security News.

    ¶¶¶¶¶

    ¶¶¶¶¶

    ¶¶¶¶¶

    ¶¶¶¶¶

    ¶¶¶¶¶

  • Anthropic has thwarted multiple sophisticated attempts by cybercriminals to misuse its Claude AI platform, according to a newly released Threat Intelligence report.

    Despite layered safeguards designed to prevent harmful outputs, malicious actors have adapted to exploit Claude’s advanced capabilities, weaponizing agentic AI to execute large-scale extortion, employment fraud, and ransomware operations.

    In one high-profile case dubbed “vibe hacking,” an extortion ring leveraged Claude Code to automate reconnaissance, credential harvesting, and network infiltration across at least 17 organizations, including healthcare providers, emergency services, and religious institutions.

    Instead of encrypting stolen data with ransomware, the group threatened to expose sensitive information to coerce ransoms exceeding $500,000 publicly.

    Claude Code autonomously selected which data to exfiltrate, determined ransom valuations based on financial records analysis, and generated alarming visual ransom notes on victim machines.

    Anthropic’s team simulated the criminal workflow for research purposes, then banned the offending accounts and developed a tailored classifier and new detection methods to flag similar behaviors in real-time.

    Another operation involved North Korean IT operatives using Claude to fabricate false identities and professional backgrounds, pass technical assessments, and secure remote positions at U.S. Fortune 500 companies.

    Where years of specialized training once throttled the regime’s capacity for such schemes, AI now enables unskilled operators to code, communicate professionally in English, and maintain lucrative employment all in violation of international sanctions.

    Upon discovery, Anthropic immediately suspended the implicated accounts, improved indicator-collection tools, and shared its findings with law enforcement and sanction-enforcement agencies.

    A third case detailed a lone cybercriminal marketing AI-generated ransomware-as-a-service on dark-web forums. Priced between $400 and $1,200 per package, the malware featured advanced evasion, encryption, and anti-recovery mechanisms, all developed with Claude’s assistance.

    Anthropic blocked the account, alerted industry partners, and enhanced its platform’s ability to detect suspicious malware uploads and code generation attempts.

    “These incidents represent an evolution in AI-assisted cybercrime,” the report warns, noting that agentic AI tools can adapt in real time to defensive measures such as malware detection systems.

    By lowering technical barriers, AI enables novices to carry out complex cyberattacks that previously required expert teams to execute. The report predicts such attacks will become more common as AI-assisted coding proliferates.

    Anthropic’s layers of protection include a Unified Harm Framework guiding policy development across physical, psychological, economic, societal, and autonomy dimensions; rigorous pre-deployment testing for safety, bias, and high-risk domains; real-time classifiers to steer or block harmful prompts; and ongoing threat-intelligence monitoring of usage patterns and external forums.

    These safeguards have already prevented misuse attempts in domains ranging from election integrity to chemical and biological weapons research, and continue to evolve in response to newly identified threats.

    In addition to account bans and detection enhancements, Anthropic has shared technical indicators and best practices with authorities and industry peers.

    Anthropic plans to prioritize further research into AI-enhanced fraud and cybercrime, expanding its threat intelligence partnerships and refining its guardrails to stay ahead of adversarial actors.

    Find this Story Interesting! Follow us on LinkedIn and X to Get More Instant Updates.

    The post Hackers Attempted to Misuse Claude AI to Launch Cyber Attacks appeared first on Cyber Security News.

    ¶¶¶¶¶

    ¶¶¶¶¶

    ¶¶¶¶¶

    ¶¶¶¶¶

    ¶¶¶¶¶

  • Anthropic, a leading AI research company, has successfully disrupted multiple attempts by cybercriminals to misuse its Claude AI model for sophisticated cyberattacks, as detailed in their latest Threat Intelligence report. The company has implemented advanced safeguards, including real-time classifiers and hierarchical summarization techniques, to detect and mitigate these abuses. By leveraging these technical measures, Anthropic […]

    The post Anthropic Blocks Hackers Attempting to Exploit Claude AI for Cyber Attacks appeared first on GBHackers Security | #1 Globally Trusted Cyber Security News Platform.

    ¶¶¶¶¶

    ¶¶¶¶¶

    ¶¶¶¶¶

    ¶¶¶¶¶

    ¶¶¶¶¶

  • It is no secret that passwords are highly susceptible to phishing and brute force attacks. This led to the mass adoption of passkeys, a passwordless authentication method leveraging cryptographic key pairs that allows users to log in with biometrics or a hardware key. According to FIDO, over 15 billion accounts have been passkey-enabled, with 69% […]

    The post Breaking the Passkey Promise: SquareX Discloses Major Passkey Vulnerability at DEF CON 33 appeared first on GBHackers Security | #1 Globally Trusted Cyber Security News Platform.

    ¶¶¶¶¶

    ¶¶¶¶¶

    ¶¶¶¶¶

    ¶¶¶¶¶

    ¶¶¶¶¶

  • The emergence of sophisticated cybercriminal organizations continues to pose significant threats to individuals and institutions worldwide, with the UTG-Q-1000 group representing one of the most concerning developments in recent cybersecurity history.

    This highly organized criminal network has demonstrated exceptional technical prowess by exploiting China’s national childcare subsidy policy, transforming what should be a beneficial government program into a vector for widespread financial fraud and data theft.

    The UTG-Q-1000 organization operates through a sophisticated multi-tiered structure, with specialized divisions including the Finance Group, News and Sex Group, Design and Manufacturing Group, and Black Market Group.

    The Finance Group specifically targets financial personnel and managers within enterprises and institutions, employing highly deceptive phishing campaigns disguised as legitimate financial communications such as tax audits, electronic receipts, and subsidy announcements.

    Their attack methodology demonstrates remarkable sophistication, utilizing multi-stage loading mechanisms through their signature “Silver Fox” remote access trojan while leveraging legitimate cloud services like Alibaba Cloud OSS and Youdao Cloud Notes to host malicious payloads and evade security detection systems.

    Qi’anxin Threat Intelligence Center researchers identified this elaborate campaign in December 2024, uncovering the group’s exploitation of the anticipated national childcare subsidy policy offering 3,600 yuan per child annually.

    The cybercriminals established numerous phishing websites overnight, mass-distributed malicious QR codes, and created convincing subsidy application pages to harvest victims’ personal information, bank card details, and authentication credentials.

    The attack infrastructure reveals a membership-based operation where individual threat actors are assigned unique identifiers to track their success rates in phishing campaigns.

    Analysis of member “ylxuqxmz” revealed 113 successful phishing attempts, with the organization maintaining detailed victim statistics across 37 compromised systems, predominantly Windows 10 machines.

    Technical Infrastructure and Evasion Mechanisms

    The UTG-Q-1000 group employs remarkably sophisticated technical evasion techniques to bypass security controls and maintain operational persistence.

    Their phishing pages function as complex loaders that dynamically create iframe containers to host the actual malicious content.

    Before loading the targeted phishing interface, the system initiates carefully disguised fetch requests to endpoints masquerading as image resources.

    The core deception mechanism involves Base64 encoding combined with XOR encryption using the key “YourSecretKey123!@#” to conceal malicious URLs within seemingly legitimate image data.

    The attack code searches for a specific signature (0x21FE) within returned image files to locate encrypted data segments, then performs the decryption process to recover target URLs and seamlessly integrate them into the victim’s browsing experience.

    async function loadContent() {
        var arrayBuffer = await_r.arrayBuffer();
        var bytes = new Uint8Array(arrayBuffer);
        for(var i=0;i<bytes.length-1;i++){
            if(bytes[i]===0x21 && bytes[i+1]===0xFE) {
                var slice = bytes.slice(i+3,l+3+l);
                var text = new TextDecoder().decode(slice);
                var url = atob(text);
                var decrypted = xorDecrypt(url, 'YourSecretKey123!@#');
            }
        }
    }

    This multi-layered obfuscation strategy effectively circumvents URL-based risk control mechanisms and static signature scanning employed by traditional security solutions.

    The organization maintains real-time victim monitoring through sophisticated heartbeat mechanisms, reporting online status every second to command and control servers at https://bmppc.cn/heartbeat.php while tracking user interactions to optimize their fraudulent operations.

    Phishing Email Interface Mimicking Official Government Communications (Source – Qi’anxin)

    The UTG-Q-1000 group represents a paradigm shift in cybercriminal sophistication, combining advanced technical capabilities with psychological manipulation to exploit public trust in government benefit programs, ultimately demonstrating the critical need for enhanced cybersecurity awareness and robust detection mechanisms.

    Boost your SOC and help your team protect your business with free top-notch threat intelligence: Request TI Lookup Premium Trial.

    The post UTG-Q-1000 Group Weaponizing Subsidy Schemes to Exfiltrate Sensitive Data appeared first on Cyber Security News.

    ¶¶¶¶¶

    ¶¶¶¶¶

    ¶¶¶¶¶

    ¶¶¶¶¶

    ¶¶¶¶¶

  • The SpiderLabs Threat Hunt Team recently discovered a cyber campaign in which threat actors used the genuine ScreenConnect remote management application as a weapon to spread the Xworm Remote Access Trojan (RAT) through a multi-phase infection chain. The attack begins with social engineering tactics, including phishing, malvertising, and deceptive social media posts, luring users to […]

    The post Weaponized ScreenConnect RMM Tool Deceives Users into Installing Xworm RAT appeared first on GBHackers Security | #1 Globally Trusted Cyber Security News Platform.

    ¶¶¶¶¶

    ¶¶¶¶¶

    ¶¶¶¶¶

    ¶¶¶¶¶

    ¶¶¶¶¶

  • Farmers Insurance has disclosed a data breach stemming from unauthorized access to a third-party vendor’s database, potentially compromising the personal information of approximately 1.1 million customers. The breach, detected on May 30, 2025, involved an unauthorized actor infiltrating a system managed by the vendor, which housed sensitive customer data. Farmers, encompassing Farmers Insurance Exchange, Farmers […]

    The post Farmers Insurance Breach Exposes Data of 1.1 Million Customers via Salesforce Compromise appeared first on GBHackers Security | #1 Globally Trusted Cyber Security News Platform.

    ¶¶¶¶¶

    ¶¶¶¶¶

    ¶¶¶¶¶

    ¶¶¶¶¶

    ¶¶¶¶¶

  • Cisco disclosed a high-severity open redirect vulnerability in the Virtual Keyboard Video Monitor (vKVM) component of its Integrated Management Controller (IMC).

    Tracked as CVE-2025-20317 with a CVSS 3.1 base score of 7.1, the vulnerability could enable an unauthenticated remote attacker to redirect administrators or users of affected devices to malicious websites, potentially capturing credentials through phishing or other social-engineering methods.

    The vulnerability stems from insufficient endpoint verification in the vKVM connection handling code. When a user clicks a specially crafted link, the vKVM client fails to properly validate the redirection target, allowing an attacker to point the user to an arbitrary URL.

    Because the IMC UI is often used for sensitive system management tasks, compromise of IMC credentials could lead to broader compromise of Cisco UCS infrastructure.

    Notably, the affected vKVM client is shared across both Cisco IMC and UCS Manager, widening the scope of devices at risk.

    Affected Products

    Any Cisco product exposing the IMC UI with a vulnerable vKVM release is impacted. Key affected platforms include:

    • UCS B-Series Blade Servers and X-Series Modular Systems.
    • UCS C-Series M6, M7, M8, and E-Series M6 Rack and Edge Servers.
    • Catalyst 8300 Series Edge uCPE.
    • All Cisco appliances built on preconfigured C-Series servers, such as APIC, DNA Center, HyperFlex, Nexus Dashboard, Secure Endpoint Private Cloud, Secure Firewall Management Center, and several others.

    Cisco has enumerated dozens of appliance families in its advisory. Only devices running fixed IMC firmware or UCS Manager software versions are safe; earlier releases remain vulnerable.

    Cisco reports that there are no workarounds available to mitigate CVE-2025-20317. Administrators must apply the security fixes released in free software updates.

    Affected customers with valid service contracts should download patched firmware via the Cisco Support and Downloads portal. Those without active contracts may contact Cisco TAC, referencing Advisory ID cisco-sa-ucs-vkvmorv-CnKrV7HK, to obtain critical fixes at no additional cost.

    Fixed Releases

    The advisory provides comprehensive tables detailing fixed firmware and software releases for each product line. Highlights include:

    • Cisco UCS Manager Software: Updates in 4.2 and above (4.2(3p), 4.3(6a)).
    • Cisco IMC on Catalyst 8300 (NFVIS): Auto-upgrade to NFVIS 4.18.1 or later.
    • UCS C-Series and E-Series Servers: Fixed IMC releases beginning with 4.2(3o) and 4.15.2, respectively.
    • Intersight-Managed Servers: Firmware 5.3(0.250001) and above for B-Series and X-Series.

    Appliance-specific remediation steps, such as applying ISO firmware updates for the Telemetry Broker or using the Cisco Host Upgrade Utility, are also outlined.

    Although Cisco has not detected any public exploitation of this vulnerability, the ease of exploitation and the sensitive nature of management interfaces make prompt remediation imperative. Organizations relying on Cisco UCS infrastructure should:

    1. Inventory all devices running Cisco IMC or UCS Manager.
    2. Determine current firmware/software versions against the advisory’s fixed-release matrix.
    3. Schedule immediate upgrades to patched versions.
    4. Review administrative procedures to ensure users avoid clicking untrusted links.

    Because stolen credentials from IMC access can facilitate lateral movement and compromise of server workloads, applying updates without delay will close this vector before attackers can weaponize it.

    CVE-2025-20317 underscores the persistent risk posed by insecure redirection in critical management interfaces. With no feasible workarounds, the sole remedy is rapid software update deployment.

    Cisco’s advisory provides the necessary guidance, and customers must act swiftly to safeguard UCS environments from credential-harvesting attacks.

    Find this Story Interesting! Follow us on LinkedIn and X to Get More Instant Updates.

    The post Cisco IMC Virtual Keyboard Video Monitor Let Attacker Direct User to Malicious Website appeared first on Cyber Security News.

    ¶¶¶¶¶

    ¶¶¶¶¶

    ¶¶¶¶¶

    ¶¶¶¶¶

    ¶¶¶¶¶

  • It is no secret that passwords are highly susceptible to phishing and brute force attacks. This led to the mass adoption of passkeys, a passwordless authentication method leveraging cryptographic key pairs that allows users to log in with biometrics or a hardware key.

    According to FIDO, over 15 billion accounts have been passkey-enabled, with 69% of users globally enabling passkeys in at least one account. The passkey promise is simple – eliminate passwords, eliminate vulnerabilities.

    Yet, SquareX researchers Shourya Pratap Singh, Daniel Seetoh, and Jonathan Lin disclosed a major passkey vulnerability at DEF CON 33 main stage that puts major banking, shopping, and enterprise SaaS app accounts at risk.

    Passkeys work by using a pair of cryptographic keys instead of a password. The private key is securely stored on the user’s device, while the public key is stored on the website’s server. When logging in, the user authenticates locally with their biometrics, local hardware key or a PIN to access the private key.

    The website then verifies this signature with the matching public key to authenticate access. This design strengthens security by tying authentication to a pre-registered device and website, eliminating the risks of stolen, reused, or weak passwords.

    Critically, all communication between the server and the user’s device is relayed through the browser. In other words, passkeys work under the assumption that the browser is “honest”.

    SquareX researchers demonstrated that through relatively trivial scripts and malicious browser extensions, attackers can intercept and forge the passkey registration process, allowing them to access accounts without the real device or biometrics.

    Even with registered passkeys, attackers can cause the passkey login to fail, forcing users to re-register their passkeys under an attacker-controlled environment.

    “Passkeys are a highly trusted form of authentication, so when users see a biometric prompt, they take that as a signal for security,” says SquareX researcher Shourya Pratap Singh, “What they don’t know is that attackers can easily fake passkey registrations and authentication by intercepting the passkey workflow in the browser. This puts pretty much every enterprise and consumer application, including critical banking and data storage apps at risk.”

    Unfortunately, traditional security tools like EDR and SASE/SSE lack the necessary visibility in the browser to detect passkey exploits. From a user perspective, the attack is identical to a legitimate passkey workflow.

    In other words, there is zero visual indicator or network signal that can verify the legitimacy of the authentication service and/or request. Thus, the only way to prevent the exploit is to monitor and block any malicious scripts and extensions directly in the browser.

    With over 80% of enterprise data now residing in SaaS applications, passkeys are emerging as the dominant authentication method for accessing these platforms.

    SquareX’s research demonstrated that browsers represent the vulnerable point in passkey security and provide the grounds for multiple attack vectors that malicious actors can leverage to exploit passkeys.

    Vivek Ramachandran, the Founder of SquareX, shares, “SquareX has been actively researching new ways attackers exploit employees in the browser.

    Without a browser security layer, passkeys in isolation can be easily hijacked by attackers to gain unauthorized access to enterprise SaaS apps, where critical data is stored. This underscores the urgent need for Browser Detection and Response, an “EDR in the browser”, which SquareX has been pioneering.”

    As passkeys establish themselves as the authentication gold standard, enterprises must ensure robust security measures are in place to protect the environment where users and passkeys primarily operate – the browser.

    About SquareX

    SquareX’s browser extension turns any browser on any device into an enterprise-grade secure browser. SquareX’s industry-first Browser Detection and Response (BDR) solution empowers organizations to proactively detect, mitigate, and threat-hunt client-side web attacks, including malicious browser extensions, advanced spearphishing, browser-native ransomware, genAI DLP, and more.

    Unlike legacy security approaches and cumbersome enterprise browsers, SquareX seamlessly integrates with users’ existing consumer browsers, ensuring enhanced security without compromising user experience or productivity.

    By delivering unparalleled visibility and control directly within the browser, SquareX enables security leaders to reduce their attack surface, gain actionable intelligence, and strengthen their enterprise cybersecurity posture against the newest threat vector – the browser.

    Users can find out more on www.sqrx.com.

    Contact

    Head of PR
    Junice Liew
    SquareX
    junice@sqrx.com

    The post Breaking the Passkey Promise: SquareX Discloses Major Passkey Vulnerability at DEF CON 33 appeared first on Cyber Security News.

    ¶¶¶¶¶

    ¶¶¶¶¶

    ¶¶¶¶¶

    ¶¶¶¶¶

    ¶¶¶¶¶

  • The China-linked advanced persistent threat (APT) actor known as Salt Typhoon has continued its attacks targeting networks across the world, including organizations in the telecommunications, government, transportation, lodging, and military infrastructure sectors. “While these actors focus on large backbone routers of major telecommunications providers, as well as provider edge (PE) and

    ¶¶¶¶¶

    ¶¶¶¶¶

    ¶¶¶¶¶

    ¶¶¶¶¶

    ¶¶¶¶¶