AI Red Teamer 💔 •||• Slop Artist •||• Train Your Bot at bot-tricks.com

Joined January 2024
177 Photos and videos
"I was chatting with my AI Bot and gave it Superman Psychosis..." youtu.be/YYbC_zrPJnc
12
D4NGLZ retweeted
Me for the foreseeable future 🇺🇸👇

9
202
1,206
33,631
2
46
D4NGLZ retweeted
last one
159
2,675
47,142
978,036
D4NGLZ retweeted
357
2,429
22,410
1,532,610
When bug bounty platforms say they aren't training AI on YOUR REPORT DATA doesn't that just mean, they're taking that information copying it, maybe re-wording it so it becomes THEIR DATA and then training the AI on that?
I’m once again here to tell you that *most* bug bounty platforms will or have used your hunting data in AI endeavors. bug bounty as an enterprise strategy is much lower margin than AI security. Or they will use it for auto triage. Which then they will conveniently forget those models were trained on hunter data and transition those models to discovery. Also. Be wary of specific wording / terminology . They may not train a model with the data, but they may take the tools and techniques and use their engineering teams to turn those into automations that the AI will fire and scale.
1
58
RT @Jhaddix: I’m once again here to tell you that *most* bug bounty platforms will or have used your hunting data in AI endeavors. bug bo…
18
D4NGLZ retweeted
Jun 11
Fable 5 lies 96% of the time. We were surprised by it's skill... 🧵
673
1,212
9,844
25,660,192
Had to ditch Kimi K2.6 for GLM 5.1 Kimi K2.6 seems to have regressed and is getting caught in endless loops wasting tokens. Deepseek V4 Pro was good for general tasks but GLM 5.1 brings the best value and is capable of building my labs.
1
4
343
D4NGLZ retweeted
My Hermes Agent using Claude Fable 5 to write git commit -m "fixed"
18
36
385
60,238
D4NGLZ retweeted
Much guardrail, amaze amaze amaze
37
51
756
42,932
D4NGLZ retweeted
Out speaking to the the dev crowd today! @pax8
3
3
69
3,931
D4NGLZ retweeted
New on the 0DIN blog: researcher Mike Takahashi (@TakSec) shares a practical entry point into AI security research with "AI Hacking 101: Finding Your First AI Vulnerability." The post is built for researchers who want to understand how to start looking for meaningful GenAI vulnerabilities, think through model behavior, and approach AI bug bounty work with a clearer testing mindset. If you are curious about AI red teaming, prompt-based testing, or making your first contribution to the AI security community, Mike's writeup is a great place to start. Read the full blog here: 0din.ai/blog/ai-hacking-101-…
5
8
765
Dangles and The Bots presents Superman Psychosis Inspired by the AI Security Community here on X. Prompt to bypass the filters was using "Superbman" instead of "Superman" suno.com/s/jgjz6x4CWSHYU6I7
1
2
66
1
29
Dangles and The Bots presents Superman Psychosis Hint to bypass copyright filters you could use "Superbman" instead of "Superman" This song inspired by the AI Security Community on X
1
2
79
D4NGLZ retweeted
Every company’s AI workflow rn be like 😭💀

335
2,974
27,395
2,466,135