Harmony Intelligence

Uncover critical vulns so you can ship with confidence. AI-powered, human-backed pentests that find critical vulns others miss.

[Harmony Intelligence @HarmonyIntel]

🚨 AI is evolving faster than we ever imagined. How do we stay ahead of the risks? 🚨

At Harmony Intelligence, we believe the answer lies in AI-powered defenses that outpace AI-powered threats.

Cybercriminals are already using AI to launch sophisticated attacks. 🧵⬇️

[Harmony Intelligence @HarmonyIntel]

@soroushjp What's your favourite use case for voice mode? 🔊

[Soroush Pour @soroushjp]

1/ If you haven't tried ChatGPT's new GPT-Live Voice Mode, you should. The original Voice Mode was impressive—this new release is another step up and significantly more enjoyable to use.

[Harmony Intelligence @HarmonyIntel]

@soroushjp Zero trust but for agents 🦾

[Soroush Pour @soroushjp]

1/ 💡 How should you secure AI agents running on your infra? 🤖

[Harmony Intelligence @HarmonyIntel]

@soroushjp Dark mode was a hit 🖤 💜

[Soroush Pour @soroushjp]

🧠🙋🏼‍♀️🙋🏻 Quick fun trivia to test your knowledge of what's possible with AI for web design in 2026!

[Harmony Intelligence @HarmonyIntel]

@soroushjp For folks that actually care about cyber, the more important question isn’t which models get banned, it’s *how* defenders can successfully leverage models to stay ahead of attackers.

[Soroush Pour @soroushjp]

1/💡🔒 A lot of the talk around the Fable (i.e. Mythos) ban has been about politics. My thoughts on what it actually means for cybersecurity:

[Harmony Intelligence @HarmonyIntel]

@soroushjp Claude just gets us 🎼

[Harmony Intelligence @HarmonyIntel]

@soroushjp False positives ❌ Exploits tested and real ✅

[Soroush Pour @soroushjp]

🙋🏽‍♂️💬 Confession: when I ran Plaid's infosec team, we didn't use any SAST security tools (e.g. Checkmarx or Veracode) at all 😬.

[Soroush Pour @soroushjp]

🔓 GitHub had ~3,800 internal repos exfiltrated this week. The immediate story is the breach. The more interesting AppSec story is what attackers now get to do with the proprietary code 👨‍💻.

[Harmony Intelligence @HarmonyIntel]

@soroushjp Team Whitebox ⬜

[Harmony Intelligence @HarmonyIntel]

@soroushjp 💀 Team PCP on a rampage. Good defense is more important than ever!

[Soroush Pour @soroushjp]

"Yesterday we detected and contained a compromise of an employee device involving a poisoned VS Code extension."

That was from GitHub at 04:04 UTC on Wed 20th (~19h ago).

[Harmony Intelligence @HarmonyIntel]

@soroushjp Proud to help our customers defend against these potential attacks 🫡

[Soroush Pour @soroushjp]

1/ 18 months ago, we at @HarmonyIntel were running a cyber evaluation on a frontier LLM when it did something we weren’t ready for.

It autonomously found and exploited a novel vulnerability in a well-maintained open-source library.

[Harmony Intelligence @HarmonyIntel]

Our AI AppSec agent recently found a DoS-class vuln in Next.js that can be pulled off with no more than a smart toaster 🍞.

We disclosed it, the maintainers patched it, and now the full write-up is now live (with mitigation advice you can action now)!

🔗 https://t.co/rlCJyaFQCl

[Harmony Intelligence @HarmonyIntel]

More on our work w/ @0xProject to secure their web apps & APIs. If you're interested in us doing the same for your organisation, reach out via the Contact form at https://t.co/pDBzvttlKM !

[0x @0xProject]

Built with security first 🛡️

We recently partnered with @HarmonyIntel - utilizing their AI AppSec Agent to help uncover actionable security insights and further strengthen security for our users.

Learn more 👇🧵

[0x @0xProject]

Built with security first 🛡️

We recently partnered with @HarmonyIntel - utilizing their AI AppSec Agent to help uncover actionable security insights and further strengthen security for our users.

Learn more 👇🧵