Check out our core values and how they shape the conference and also our organizer team: opensecurityconference.org/about/values/
#osco #osco26 #Security #CyberSecurity #InfoSec #AppSec #OTSecurity [lisi] 2/2

Check out our core values and how they shape the conference and also our organizer team: opensecurityconference.org/about/values/
#osco #osco26 #Security #CyberSecurity #InfoSec #AppSec #OTSecurity [lisi] 2/2
The latest update for #AppSentinels includes "What Google Gemini's Sandbox Escape Reveals About Securing #APIs Against #AI Agents" and "What Does #NIST IR 8587 Mean for #API Security?".
#cybersecurity #APISecurity #AppSec https://opsmtrs.com/3MZSK4Z
#OWASP #Ottawa is pleased to announce that Alexander Rudolph, PhD presenting @ OWASP Ottawa Day 2026 on:
"Canadian Digital Sovereignty in a Zero-Trust World"
📅 October 17, 2026
ℹ️ Information: github.com/OWASP-Ottawa...
#DigitalSovereignty #CyberSecurity #Canada #OWASPOttawa2026 #AppSec
On behalf of #OWASP #Ottawa, a huge thank you to Mark Cimon and the Cimon Real Estate Team for supporting OWASP’s 25th Anniversary event at OWASP Ottawa Day!
Information and tickets:
github.com/OWASP-Ottawa...
#OWASP #Ottawa #OWASPOttawa2026 #AppSec #Cybersecurity #Community
💰 $30–$100/HR | APPLICATION SECURITY ENGINEER 🔐💻
🌍 Remote (~15 hrs/week)
Experienced in AppSec & backend?
Use your expertise to help train next-generation AI.
🔐 Pentesting • OWASP • Secure Coding • Python • Java
CSP Nonces and the Same-Origin Policy Solve Different Problems www.darinor.com/blog/csp-non... #Appsec #WEBSecurity #CSP
Security Tip: Implement a Software Bill of Materials (SBOM) strategy. 🛡️ Knowing exactly what's in your software stack is the first step to securing it. Use tools to automate SBOM generation and track component risks. Learn more: https://cvedatabase.com #SBOM #InfoSec #AppSec
When people say PHP isn't secure, I point them to how every release improves security a little bit. PHP 8.6 makes sessions secure by default (check your SSO and payment callbacks before upgrading).
https://www.phparch.com/2026/09/stop-writing-php-like-this-8
🎉 Last call! Offer ending September 30th! 🎁 Join OWASP between today and 30th Sept and get 1 FREE year's individual membership to gift to a friend. More people, more knowledge, more AppSec! 🔐
One membership. Twice the impact. bit.ly/OWASPBringaFr...
300 lines of Python that fuzzes web templates for SSTI and flags vulnerable engines. Advanced. Run: python3 ssti_tester.py --url URL
https://www.valtersit.com/python/server-side-template-injection-ssti-tester/
#security #python #appsec
The security question for agent tools is not just 'does it work?' It is 'what can it do when it is wrong, compromised, or over-permissioned?' #AIAgents #AppSec
Lesson: Never ship static credentials in client bundles.
Read the full engineering post-mortem on Dev.to:
dev.to/freerave/the...
Who says CSRF isn't a real vulnerability?
https://thehackernews.com/2026/09/elementor-csrf-flaw-lets-attackers-take.html
Burned out development teams and people sleeping during threat modeling sessions.
The realities of doing security is not that pink as we want to portray it, and our tools and processes not that smooth.
#cornucopia #security #awarness #games #threatmodeling #threatmodelling #appsec
The latest update for #GitGuardian includes "#AI Coding Agents Are Leaking Credentials: Cursor, Claude Code, Copilot, and MCP" and "How the GitGuardian Mixin Kit Extends #Docker Sandboxes for Safer AI Coding".
#cybersecurity #DevOps #infosec #appsec https://opsmtrs.com/3XY1xZb
Supply chain attacks aren't just about injecting malicious code—they're credential theft operations targeting developer machines and CI/CD pipelines. Here's what defenders are missing.
https://blog.gitguardian.com/supply-chain-attack-targets-credentials/
474 valid GitHub App private keys leaked in the wild—some with org admin access. GitGuardian's research shows the scale of secrets exposure risk and why secrets detection matters for DevSecOps…
https://blog.gitguardian.com/github-app-private-keys-leaked/
Three major AI agent vendors just got caught trusting the pin, not what's pinned. Plugin4Shell shows how attackers can swap malicious code into Copilot, Claude, and Gemini—a supply…
https://dev.to/humanbound_ai/three-ai-agents-one-trust-boundary-problem-14l
CISA's 2026 SBOM rules mandate component hashes—but one security engineer argues they're the weakest link. Here's why verification by hash alone may not be enough.
https://www.reddit.com/r/devsecops/comments/1wg5jps/component_hashes_in_sboms/