• ChatGPT Image Outage Causes Generation Errors and Upload Delays

    OpenAI is investigating an ongoing ChatGPT image outage that is disrupting image generation and file uploads. Users report that ChatGPT freezes when they upload files. Others encounter errors after submitting image-generation prompts. The incident also affects developers using the Images API. Users report image generation failures The problems began more than an hour before OpenAI…

  • SAP Patches Maximum-Severity OVERPASS Kernel Vulnerability

    SAP has released security updates for 20 vulnerabilities across several products. The patches include a maximum-severity SAP kernel vulnerability that could allow attackers to take full control of affected systems. The critical flaw, tracked as CVE-2026-44756, affects the Extended Passport Protocol processing library. Security researchers at Onapsis discovered the issue and named it OVERPASS. OVERPASS…

  • Hackers Build AI Frameworks for Widescale Credential Theftg

    Cybercriminals are moving beyond simple AI coding tools. Instead, they are building multi-agent systems that can automate large parts of an attack with little human supervision. Google’s Threat Intelligence Group observed AI agents coordinating attacks, solving technical problems, and adapting their actions in real time. Some systems also scanned for vulnerabilities and collected thousands of…

  • Vietnam-Linked APIS Data Leak Exposes 220 Million Travel Records

    An exposed passenger information database revealed more than 220 million travel records linked to passengers and airline crews. The APIS data leak included passport details, flight information and other sensitive personal data collected over nine years. Researchers traced the system to an IP address in Hanoi, although they could not identify the Vietnamese organization responsible…

  • Magento StyleSmuggler zero-day deploys Linux backdoor

    Attackers are exploiting a zero-day vulnerability affecting all versions of Magento and Adobe Commerce. The flaw, named StyleSmuggler, allows hackers to execute malicious code and install a persistent Linux backdoor. E-commerce security company Sansec detected the first known attack on September 4. The targeted website had already installed the latest available security updates. Adobe Enterprise…

  • BigBear phishing service bypassed MFA at 258 organizations

    A phishing-as-a-service platform called BigBear 2.0 has compromised Microsoft 365 accounts at 258 organizations. The operation collected more than 5,000 credential records by intercepting passwords and authenticated session cookies. CloudSEK researchers gained administrator access to the service’s control panel. Their investigation uncovered 42 virtual private server nodes configured to target Microsoft 365 users. The BigBear…

  • Mathspace data breach affects over 1 million people

    Online learning platform Mathspace has disclosed a major data breach affecting more than one million students, parents, guardians and staff members. Attackers stole the information after compromising the company’s internal reporting system. The incident affected 1,079,819 people in Australia and New Zealand. However, Mathspace said the attackers did not obtain passwords, authentication credentials or academic…

  • Trezor data breach now affects 81,000 customers

    Cryptocurrency hardware wallet maker Trezor says a breach at shipping provider ShipMonk has exposed information belonging to 81,000 customers. The total increased after the company discovered that another 67,000 US customers were affected. Trezor initially reported the incident on August 13. At that point, the company believed attackers had accessed the personal information of almost…

  • Hackers exploit MikroTik RouterOS flaws to hijack routers

    Hackers are actively exploiting two critical MikroTik RouterOS flaws to take control of internet-exposed routers. The attack chain targets devices that make their SSH services accessible through public networks. Poland’s CERT agency discovered the vulnerabilities with assistance from GPT-5.5-cyber and GPT-5.6-sol. Researchers named the exploit chain “MikroTrick” and confirmed that attackers are already using it…

  • ConnectWise warns of ScreenConnect flaw with no patch available

    ConnectWise has disclosed a new vulnerability affecting file transfers in its ScreenConnect Remote Access platform. A permanent patch is not yet available, so the company has issued temporary mitigation instructions for customers. The ScreenConnect flaw affects both cloud-hosted and on-premises deployments. However, ConnectWise has not yet assigned it a CVE identifier or disclosed detailed technical…