Your AI Agent Has Admin Access and You Don't Know It

This story was originally published on HackerNoon at: https://hackernoon.com/one-identity-unveils-identity-manager-100-lts-with-ai-and-modern-ux. Of the three products, Identity Manager 10.0 LTS provides the clearest UI modernization story. Its new browser-based experience provides administrative function Check more stories related to cybersecurity at: https://hackernoon.com/c/cybersecurity. You can also check exclusive content about #cybersecurity, #cyber-threats, #cyber-security-awareness, #cybernewswire, #cybercrime, #cyberthreats, #cybersecurity-tips, #good-company, and more. This story was written by: @cybernewswire. Learn more about this writer by checking @cybernewswire's about page, and for more stories, please visit hackernoon.com. One Identity unveiled Identity Manager 10.0 LTS, modernizing the identity governance experience with browser-based administration, AI-assisted insights, and streamlined automation. The new release provides a unified cross-platform experience, accessible synchronization, and AI-assisted reporting, making complex governance more accessible.

This story was originally published on HackerNoon at: https://hackernoon.com/offsec-celebrates-20-years-of-practitioner-built-training-from-the-creators-of-kali-linux-and-oscp. This month OffSec is marking its 20th anniversary as a global leader in hands-on cybersecurity training and certifications for individuals and Check more stories related to cybersecurity at: https://hackernoon.com/c/cybersecurity. You can also check exclusive content about #cybersecurity, #cyber-threats, #cybernewswire, #press-release, #cyber-security-awareness, #cybercrime, #cyber-attack, #good-company, and more. This story was written by: @cybernewswire. Learn more about this writer by checking @cybernewswire's about page, and for more stories, please visit hackernoon.com. OffSec, a global leader in hands-on cybersecurity training and certifications, is marking its 20th anniversary with a renewed focus on continuous skill development and AI-era learning. The company's mission remains to develop critical thinkers, cyber warriors, and problem solvers who are ready to excel in their jobs under stress.

This story was originally published on HackerNoon at: https://hackernoon.com/insignary-releases-free-insignary-chk-tool-to-detect-php2shell-and-node2shell-exposure. Node2Shell (CVE-2026-111497) is an out-of-bounds write flaw in Google’s Brotli decompression library, which Node.js bundles. Specially crafted Check more stories related to cybersecurity at: https://hackernoon.com/c/cybersecurity. You can also check exclusive content about #cybersecurity, #cybercrime, #cybernewswire, #cyber-threats, #cyber-attack, #cybersecurity-tips, #cyber-resilience, #good-company, and more. This story was written by: @cybernewswire. Learn more about this writer by checking @cybernewswire's about page, and for more stories, please visit hackernoon.com. Insignary released insignary-chk, a free tool that checks for PHP2Shell and Node2Shell, two recently disclosed runtime vulnerabilities affecting PHP and Node.js. The tool is a single file that requires no installation or registration and sends no data over the network unless the user chooses to fix the issues.

This story was originally published on HackerNoon at: https://hackernoon.com/your-ai-agent-has-admin-access-and-you-dont-know-it. Your AI agent has admin access and you don't know it. How demo credentials become permanent attack surface, and the permission audit that fixes it. Check more stories related to cybersecurity at: https://hackernoon.com/c/cybersecurity. You can also check exclusive content about #cybersecurity, #ai, #ml, #security, #software-engineering, #ai-agents, #production-agents, #ai-agents-permissions, and more. This story was written by: @kishoreyarram. Learn more about this writer by checking @kishoreyarram's about page, and for more stories, please visit hackernoon.com. Most teams can only list three items of what their AI agent can touch, but in reality, the agent has a broader blast radius due to unscoped permissions and integrations. The fix is to apply the discipline of least privilege to agents, giving them narrow, auditable interfaces and standard protocols.

This story was originally published on HackerNoon at: https://hackernoon.com/will-open-source-software-survive-the-ai-vulnerability-flood. AI vulnerability discovery could make open source safer while overwhelming maintainers, exposing systemic dependencies, and reshaping the software ecosystem. Check more stories related to cybersecurity at: https://hackernoon.com/c/cybersecurity. You can also check exclusive content about #ai-security, #software-security, #ai-vulnerability-disclosure, #vulnerability-management, #software-supply-chain, #devsecops, #open-source-ai-security, #cyber-resilience-act, and more. This story was written by: @sebastianmartinez. Learn more about this writer by checking @sebastianmartinez's about page, and for more stories, please visit hackernoon.com. AI-assisted vulnerability discovery could trigger a temporary flood of reported vulnerabilities as decades of hidden software flaws become easier to find. Open source may experience this surge earlier and more visibly than closed source, but higher CVE counts do not necessarily mean weaker security. With effective remediation and continuous AI-assisted security analysis, surviving projects could become more secure. However, accelerating discovery, regulatory pressures, and longstanding dependence on underfunded maintainers may threaten critical projects, force ecosystem reorganization, and concentrate systemic risk.

This story was originally published on HackerNoon at: https://hackernoon.com/fake-google-update-extension-turns-vs-code-into-a-malware-delivery-tool. How a fake Google Update Support extension installs its components and runs the editor's own Electron binary as a Node interpreter. Check more stories related to cybersecurity at: https://hackernoon.com/c/cybersecurity. You can also check exclusive content about #cybersecurity, #malware, #vscode, #electron, #threat-intelligence, #developer-security, #social-engineering, #javascript, and more. This story was written by: @ilja-krasnjanski. Learn more about this writer by checking @ilja-krasnjanski's about page, and for more stories, please visit hackernoon.com. A take-home repository from a fake job interview starts a VS Code task when the folder is opened, installing an extension called Google Update Support. The 17.6 MB package carries native agents for macOS and Linux, encrypted Windows components, and a path that runs the signed editor binary as a plain Node interpreter. Reconstructed from the artifacts without running any sample.

This story was originally published on HackerNoon at: https://hackernoon.com/binances-post-mortem-is-out-the-$63-billion-hack-started-with-a-job-offer. Binance's post-mortem traces the $6.3B hack to a fake job interview, a tampered install and an internal AI tool. Here's the five-month timeline. Check more stories related to cybersecurity at: https://hackernoon.com/c/cybersecurity. You can also check exclusive content about #cybersecurity, #binance, #eip-7702, #binance-hack-6.3-billion, #binance-hack-post-mortem, #binance-hack-socialengineering, #binance-hack-report, #cybersecurity-top-story, and more. This story was written by: @sidrazafar. Learn more about this writer by checking @sidrazafar's about page, and for more stories, please visit hackernoon.com. Binance's post-mortem shows the record theft began months earlier: a tampered install in June, a fake coding test in September, an internal AI tool in October. Then two hours and 41 minutes inside the wallet system, and a security design that made the cleanup slower.

This story was originally published on HackerNoon at: https://hackernoon.com/the-sandbox-fights-back-joe-security-launches-adaptive-detonation. AI agents help Joe Sandbox uncover malicious behavior concealed by evasion techniques and unmet execution conditions. Check more stories related to cybersecurity at: https://hackernoon.com/c/cybersecurity. You can also check exclusive content about #cybersecurity, #cyber-threats, #cybernewswire, #press-release, #cyber-security-awareness, #cybersecurity-tips, #cyber-attack, #good-company, and more. This story was written by: @cybernewswire. Learn more about this writer by checking @cybernewswire's about page, and for more stories, please visit hackernoon.com. Joe Security's Adaptive Detonation uses AI agents to intervene in malware analysis, helping security teams uncover hidden behavior. This approach provides deeper visibility into evasive threats and more evidence to support investigations.

This story was originally published on HackerNoon at: https://hackernoon.com/your-startups-biggest-security-vulnerability-isnt-your-code-its-you. Hackers don't break in. They log in. Here's the founder's field manual for not handing them the keys. Check more stories related to cybersecurity at: https://hackernoon.com/c/cybersecurity. You can also check exclusive content about #cybersecurity, #startups, #founders, #security, #small-business, #cyber-security-awareness, #security-breaches, #cybersecurity-tips, and more. This story was written by: @therealmdwhite. Learn more about this writer by checking @therealmdwhite's about page, and for more stories, please visit hackernoon.com. I’ve spent 25+ years in business warfare — buying troubled companies, turning them around, doing private security consulting, and researching cybersecurity from the operator’s side, not the engineer’s. And I can tell you this with total confidence: when small businesses get hacked, it’s almost never some hoodie-wearing genius cracking encryption. It’s a founder reusing a password. It’s an admin clicking a link. It’s someone handing over the keys because the “IT guy” on the phone sounded convincing.

This story was originally published on HackerNoon at: https://hackernoon.com/the-mcp-server-chose-the-token-endpoint-two-official-sdks-trusted-it. I found OAuth credential exposure flaws in the official MCP Python and TypeScript SDKs. Why patching CVE-2026-115563 and CVE-2026-104850 may not be enough. Check more stories related to cybersecurity at: https://hackernoon.com/c/cybersecurity. You can also check exclusive content about #cybersecurity, #ai-security, #cve, #programming, #security-vulnerabilities, #token-endpoints, #mcp-server, #sdk, and more. This story was written by: @aviralxroot. Learn more about this writer by checking @aviralxroot's about page, and for more stories, please visit hackernoon.com. I reported two OAuth credential exposure flaws in the official MCP Python and TypeScript SDKs, now tracked as CVE-2026-115563 and CVE-2026-104850. A malicious MCP server could manipulate OAuth discovery, causing affected clients to send secrets, refresh tokens, or other credentials to attacker-controlled endpoints. Both SDKs have patches, but some integrations also need explicit issuer binding and updates to stored credentials.