Click any tag below to further narrow down your results
Links
The author argues that the surge in security incidents and AI-written code calls for engineering fixes rather than more AI, and walks through using Nix to patch the whole stack from one source of truth. The core idea is that agents make open-source dependencies less necessary, and that Nix overlays let you route around upstream bugs without waiting for maintainers.
- Agent-written code trained on existing vulnerable code is driving a "security take-off," and defenders may need more inference capacity than attackers to keep up.
- Open-source libraries existed because code was costly to write; with agents, the author uses far fewer and prompts for small, purpose-built code instead, reducing Dependabot and supply-chain churn.
- Nix can act as a single source of truth for laptops, CI, agent sandboxes and production images, and lets you monkey-patch tools like git (removing force push) across every environment at once.
This daily roundup covers Fortinet’s FortiBleed campaign exposing 86,000 device credentials, a Texas hunting-license vendor breach affecting 3 million records, and an unpatchable BootROM exploit on Apple A12/A13 chips. It also highlights GitHub’s context-aware secret scanning, the Novo Nordisk code leak via a stolen GitHub token, and other emerging tools and vulnerabilities.
- FortiBleed exposed 86,000+ valid Fortinet credentials across 194 countries via 1.16 billion login attempts against 320,000 FortiGate devices, with hashes cracked on a 45-GPU cluster.
- A Texas hunting/fishing license vendor breach leaked driver's license and passport numbers for 3.09 million people, already being sold by a threat actor linked to a prior Virginia wildlife breach.
- The unpatchable "usbliter8" BootROM exploit affects Apple A12/A13 chips (iPhone XS to iPhone 11), installing a restart-persistent handler that downgrades security and boots unsigned code.
- GitHub's new context-aware LLM secret scanning cuts false positives by over 75% by isolating code paths where secrets feed into API calls, auth headers, or database clients.
- FulcrumSec stole a GitHub PAT to exfiltrate 1.3 TB from Novo Nordisk, including the Ozempic formula and clinical-trial data for 11,500 patients, leaking 264 GB after a refused $25 million ransom.
This daily roundup covers a 40 GB data breach at the University of Nottingham, a lost-drive incident exposing 10.9 million Japanese utility customers, and a proof-of-concept Exchange spoofing flaw. It also highlights automated AI-driven attack research, supply-chain toolkits on GitHub, and new product launches for dependency patching and taint analysis.
- ShinyHunters exploited a zero-day gadget chain in University of Nottingham's PeopleSoft to steal 40GB of data on 454,600 students, including passport numbers and disability details.
- Kyushu Electric Power lost a physical backup drive exposing personal and usage data for 10.9 million customers.
- InfoGuard's "Ghost-Sender" exploit spoofs any Exchange Online address (even CEO/noreply) via a one-line PowerShell script, bypassing SPF, DKIM, and DMARC.
- A researcher earned over $500,000 in bug bounties using an AI-driven fuzzer to scrape thousands of Google API keys and find leaky internal-only endpoints.