Application Security News and Articles


Over-scoped agents: The permission sprawl that will end you

Every engineering team is spinning up agents. And they're all over-scoped because nobody wants to be the person whose agent doesn't work. The post Over-scoped agents: The permission sprawl that will end you appeared first on Strata.io. The post ...

Blind spots: Your agents are operating in complete darkness (and so are you)

Traditional IAM logs were built for humans clicking buttons. Your agents are executing complex decision trees at machine speed. It's like trying to track Formula 1 with a sundial. The post Blind spots: Your agents are operating in complete ...

Human-in-the-loop: When your AI’s creativity becomes your legal liability

Here's the truth nobody wants to admit: Your autonomous agents are making promises right now that you don't know about, can't afford, and will be legally forced to honor. The post Human-in-the-loop: When your AI’s creativity becomes your legal ...

Agent credential replay: Why bearer tokens are digital cash in a tornado

Here's what should keep you up at night: Your agents are passing these digital IOUs to each other thousands of times per second. Each handoff is an opportunity for theft, confusion, or replay. The post Agent credential replay: Why bearer tokens ...

Randall Munroe’s XKCD ‘’Piercing”

via the comic artistry and dry wit of Randall Munroe, creator of XKCD Permalink The post Randall Munroe’s XKCD ‘’Piercing” appeared first on Security Boulevard.

It’s a Partnership, Not a Handoff: Doug McLaughlin on Navigating Enterprise Change

The journey from a signed contract to a fully deployed security solution is one of the most challenging in enterprise technology. For a mission-critical function like identity, the stakes are even higher. It requires more than just great ...

USENIX 2025: PEPR ’25 – Using GenAI to Accelerate Privacy Implementations

Creators, Authors and Presenters: Rituraj Kirti, Software Engineer - Meta & Rajkishan Gunasekaran, Software Engineer - Meta Our thanks to USENIX for publishing their Presenter’s outstanding USENIX Enigma ’23 Conference content on the ...

The Psychology of Security: Why Users Resist Better Authentication

70% of Americans feel overwhelmed by passwords, yet only half choose secure ones despite knowing the risks. The problem isn't user education—it's psychology. Discover why users resist better authentication and the UX design principles that make ...

In Other News: Gladinet Flaw Exploitation, Attacks on ICS Honeypot, ClayRat Spyware

Other noteworthy stories that might have slipped under the radar: US universities targeted by payroll pirates, Zimbra vulnerability exploited, Mic-E-Mouse attack. The post In Other News: Gladinet Flaw Exploitation, Attacks on ICS Honeypot, ...

OpenVPN redefines secure connectivity with Access Server 3.0

OpenVPN released Access Server 3.0, a major update to its self-hosted business VPN solution that delivers foundational improvements to performance, flexibility, and system integration. While the most visible change is a modernized Admin Web UI, ...

Accenture helps organizations advance agentic AI with Gemini Enterprise

Accenture and Google Cloud announced that their strategic alliance is driving client reinvention with Gemini Enterprise agentic AI solutions, building on the successful adoption of Google Cloud technologies for organizations across industries. ...

Proof launches Certify, the cryptographic answer to AI-generated fraud

Generative AI is enabling the proliferation of fake documents, images, videos, and data at an unprecedented scale, to the point where it’s indistinguishable from reality. While fake media and misinformation have garnered the most attention, the ...

comforte AG debuts TAMUNIO, its all-in-one shield for data security

comforte AG launched TAMUNIO, a unified data security platform designed to reduce risk, accelerate innovation with cloud and AI, and optimize operational costs for the most demanding enterprises. Built on decades of experience securing ...

Apple offers $2 million for zero-click exploit chains

Apple bug bounty program’s categories are expanding and rewards are rising, and zero-click exploit chains may now earn researchers up to $2 million. “Our bonus system, providing additional rewards for Lockdown Mode bypasses and ...

Mobilicom rolls out a secured autonomy system powering next-gen AI drones

Mobilicom launched the Secured Autonomy (SA) Compute PRO-AT, which combines Mobilicom’s OS3 (Operational Security, Safety, and Standards compliance) cybersecurity software with Aitech’s rugged, NVIDIA-based AI Supercomputers, including the ...

Cybersecurity Snapshot: AI Security Skills Drive Up Cyber Salaries, as Cyber Teams Grow Arsenal of AI Tools, Reports Find

Want recruiters to show you the money? A new report says AI skills are your golden ticket. Plus, cyber teams are all in on AI, including agentic AI tools. Oh, and please patch a nasty Oracle zero-day bug ASAP. And get the latest on vulnerability ...

Cisco, Fortinet, Palo Alto Networks Devices Targeted in Coordinated Campaign

GreyNoise has discovered that attacks exploiting Cisco, Fortinet, and Palo Alto Networks vulnerabilities are launched from the same infrastructure. The post Cisco, Fortinet, Palo Alto Networks Devices Targeted in Coordinated Campaign appeared ...

What is CAA? Understanding Certificate Authority Authorization

Learn what a CAA record is, how it protects your domain from unauthorized SSL certificate issuance, and how to set it up for stronger website security. The post What is CAA? Understanding Certificate Authority Authorization appeared first on ...

RondoDox Botnet Takes ‘Exploit Shotgun’ Approach

The botnet packs over 50 exploits targeting unpatched routers, DVRs, NVRs, CCTV systems, servers, and other network devices. The post RondoDox Botnet Takes ‘Exploit Shotgun’ Approach appeared first on SecurityWeek.

Multimodal AI, A Whole New Social Engineering Playground for Hackers

Multimodal AI delivers context-rich automation but also multiplies cyber risk. Hidden prompts, poisoned pixels, and cross-modal exploits can corrupt entire pipelines. Discover how attackers manipulate multimodal inputs—and the governance, ...