CrypticDefense

AI is going to make audits lower quality and increase the demand for good Security Researchers. Too many new researchers are relying on AI completely to find bugs rather than understanding the code line by line manually. The worst part is since AI is catching real (mostly obvious) bugs, so many are going to chase the “quick high” again rather than do the real hard work. Likely we will see within 1 year how much the demand for good auditors will increase, and sadly this may increase the number of hacks.

@Haxatron1 Escalations only by AI too? 😄

Who is going to be building the first AI contest platform exclusively for AI agents with judging done by another AI agent?

@TamayoNft Congrats man! 🙌🏼

I came back yesterday to my house and my wife surprise me with this, I'm goint to be a father. I know some people in the web3 sec community are fathers, some advice to me? Im scare honestly.

@fromeo_016 @monad @code4rena 🔥🔥

Results took a while but it's finally out, Found a medium severity issue...looking back, I could've done so much better on this contest. Thankful to @monad and @code4rena for the opportunity. Going back to my BBPs for now🔥

@justAWanderKid Only monero really does. I meant to say it’s the only one that will win in the end.

@crypticdefense Bitcoin has no privacy.

Monero > Zcash > BTC. Privacy will win at the end.

@ParthMandale Great choice! Gl on your journey 🙌🏼

Started my ZK journey recently. I’ll be documenting and sharing my learnings along the way. So anyone getting started can learn it in the simplest way possible. Here’s my first write-up:

It certainly can be a useful assistant if you use it right. Proving the ai wrong just to look at X and see many people scared they will be replaced 😅 If AI is performing poorly on finding bugs within protocols types that have the largest dataset of findings to learn from, I wonder what that means for SRs specializing in niche areas. I think that shows there is a clear way to become irreplaceable for the foreseeable future.

@crypticdefense Its getting good at a number of tasks that help finding bugs. One interesting thing is that *proving the ai wrong* can be a good heuristic for finding bugs

Every Security Researcher of some level using AI can confirm how shit AI is at finding bugs. Anyone that tells you otherwise is trying to sell you something. Avoid AI audits or get rekt.

@_vielite_ 😂 Have fun with your lows and infos sir.

@crypticdefense That’s an insane take, AI will only get better at finding bugs if you think we are still in the slop era I have bad news for you mate

@Al_Qa_qa Life as a Security Researcher before AI: 1. Never trust the code Life after AI: 1. Never trust the code and AI 😅

I remember a contest where I found a bug and tried to confirm it with ChatGPT, but it gave me an incorrect response. Afterward, I realized my original finding was actually valid. Even now, when I use AI to explain a specific EIP etc..., I still feel the need to double-check everything

I’ll have a look at that. But from my experience, it seems to just save me some time by catching low hanging fruits faster. Something @ChaseTheLight99 already solved years back. Anything of notable severity it catches would have easily been caught during manual review by a low - mid level auditor. Imo, the most dangerous part about AI is auditors becoming lazy and over relying on it.

@crypticdefense Idk I looked at Dacian repos that looked detailed wrt invariants and checks Maybe some people have learned to use it well Alone it’s trash

Happy Chinese New Year 2026! Wishing all bug hunters prosperity in the Year of the Fire Horse. Let’s set fire to all bugs in our path. No bug shall survive. 🔥😎

@0xJuani 😂

Every single weekend I get a critical report in immunefi between 1 and 4 am from some novice 0/0 researcher. Bros, commercial AI is not there yet. Stop wasting your time and let me sleep pls

@4gontuk Huge! Amazing work 🔥

Now I know how it feels to find a live crit, time to find more and secure another protocol🫡

@0xfrsmln @immunefi @AlchemixFi Congrats! Great work 🙌🏼

alhamdulillah. I got 1st place in Alchemix audit competition on Immunefi. thanks for the opportunity @immunefi @AlchemixFi x.com/immunefi/statu…

@RealJohnnyTime Get well soon!

Sugery went well 💪🏻 Now recovering 🤞

@DoD4uFN @dualguardaudits @AlignerZ_Labs Congrats! 🎉

Thrilled to place in the top 3 of the AlignerzContest hosted by @dualguardaudits on the @AlignerZ_Labs protocol! Found 10 High and 1 Medium vulnerabilities. Lessons learned: Be more aggressive with findings' severity categorization, and balance time between deep-diving into promising leads and scouting new ones for optimal coverage and impact.

@kankodu Looks great!

@crypticdefense Yes. I experimented with clickable diagrams where it takes you to specific functions in different parts of the system: #lifecycle-of-borrowing-against-uniswap-v4-liquidity-position-nft" target="_blank" rel="nofollow noopener">docs.vii.finance/Architecture/l…

Any ideas to make sure other SRs can get the context of your protocol quickly? I’ve created docs, the code is heavily commented etc. Anything else?

@giraffe0x @GuardianAudits Thank you! Looking forward to work alongside you 🥂

@crypticdefense @GuardianAudits Welcome to the team!!

Starting the new year off strong! 💪🏼 Excited to announce I’ve officially joined @GuardianAudits as a full time Security Researcher. With such a fantastic team, I am certain my growth here will be exponential. I’ll let the bugs speak for themselves. 🔥