Aikido
Start for Free
No CC required

Debarshi

Researcher
Debarshi is an AI researcher at Aikido. Previously, he was a member of the technical staff at Hacktron AI. He stays late to benchmark the latest AI models and regularly posts on X about what he finds.
AI Benchmarking
Aikido Libraries
Data Security
Hardened images
GitHub Security
Mythos
Aikido Device Protection
Aikido Endpoint
Pypi
Secrets
Aikido Zen
IDOR vulnerabilities
IDE Security
Announcements
European Cyber Security
Continuous Pentesting
AI Safety
Self-securing Software
SSDLC
VS Code
AI Penetration Testing
Threat Modeling
Cyber Resilience Act
Triaging
AutoTriage
Pentesting
Bazel
Code Quality
OWASP
NIS2
Legaltech
fintech
RASP
End of Life
SQL Injections
DAST
cnapp
CSPM
ASPM
Infrastructure as a Code IaC
Network Security Monitoring
License Scanners
SBOM
Container Scanning
AppSec
Vulnerabilities
Software Supply Chain Security
API
Dependencies
Continuous Security Monitoring
DevSecOps
Vibe Coding
AI
NPM
autofix
Malware
Article
open-source
SCA
intel
SOC 2
usecase
Tools
SAST
Compliance
CircleCI
Codeship
IMDSv2
Cloud
IMDSv1
SSRF
AWS

Blog posts by Debarshi

August 24, 2026
•
Research

How Aikido finds more vulnerabilities than Mythos at half the cost

Aikido AI Code Audit found 8 more vulnerabilities than Claude Security with Mythos at less than half the cost. How harness design drives coverage per dollar.

Tags/

#AI Penetration Testing

#AI

#Mythos

August 21, 2026
•
Research

We burned 11.7bn tokens to find the best cyber AI model

We tested 10 AI models on 32 fresh CVEs. DeepSeek V4 Pro found 28, and three cheap runs beat one pass of Opus 5 or Grok on total coverage.

Tags/

#AI

#AI Penetration Testing

#AI Benchmarking

Get secure now

Secure your code, cloud, and runtime in one central system.
Find and fix vulnerabilities fast automatically.

Start Scanning
Book a demo
No credit card required | Scan results in 32secs.