Wyatt @whyitfor

Benevolent binary manipulator at @redballoonsec. OFRAK! ofrak.com NYC Joined July 2022

Tweets

149
Followers

39
Following

248
Likes

90

Red Balloon Security @redballoonsec

2 months ago

This is RASPUTIN—the first human-on-the-loop hardware-reversing robot, developed by Red Balloon Security. Automate embedded hardware analysis, firmware extraction, vulnerability detection, fault testing, and more. 🔗 Learn more: redballoonsecurity.com/industry-solut… #Firmware #Engineering

0 4 7 1K 5

Download Video

Thomas H. Ptacek @tqbf

3 months ago

People keep making category errors talking about how insidiously broken LLM-generated code is. The broken assumption: developers are on the hook to accept any generated code. It's weird, b/c I reject about 1/3rd of all generated PRs outright.

11 6 144 10K 24

Red Balloon Security @redballoonsec

4 months ago

It was great meeting everyone who visited our booth at #EScarUSA! We showcased our latest research and solutions on securing embedded automotive ECUs and joined the conversation on the #EMB3D threat model. #AutomotiveSecurity #CyberSecurity #ECU #EmbeddedSystems

0 1 1 180 0

Download Image

Brian in Pittsburgh @arekfurt

5 months ago

Oh come the f***🤬 on. You gotta be kidding. If an attacker can get a user to click a bloody WebEx meeting invite link they can compromise the user's device. Because version 44.6 of WebEx stopped validating meeting invite links content properly. sec.cloudapps.cisco.com/security/cente…

5 84 290 32K 93

Download Image

Wyatt @whyitfor

5 months ago

Pretty interesting phish

nick.eth @nicksdjohnson

5 months ago

Pretty interesting phish

1K 6K 36K 5.7M 17K

Download Image

0 0 0 20 0

Jake Williams @MalwareJake

5 months ago

If you work in DevOps read this post-mortem. If you don't work in DevOps, send this to your coworkers that do. Trust me, it's important they read this. landh.tech/blog/20250211-…

6 58 254 23K 281

Wyatt @whyitfor

6 months ago

😾😾😾

Ryan Hurst @rmhrisk

6 months ago

😾😾😾

0 6 5 976 3

0 0 0 42 0

Wyatt @whyitfor

6 months ago

Adjacent experience—we asked an LLM to implement a parsing function and cite source. Its source was a blog post which lacked an input validation, but LLM added this check silently. When asked where it got the validation from, LLM said it read the RFC:)

AIfredo 0rtega @ortegaalfredo

6 months ago

8 9 28 11K 6

0 0 0 41 0

Alex Matrosov @matrosov

7 months ago

Nice writeup on how legacy bugs still work in new AI reality. Ultimately, we are still dealing with a lot of C code, and the same problems appear again.

Ruikai @retr0reg

7 months ago

Nice writeup on how legacy bugs still work in new AI reality. Ultimately, we are still dealing with a lot of C code, and the same problems appear again.

4 106 416 40K 233

1 8 28 6K 9

Dino A. Dai Zovi @dinodaizovi

7 months ago

It's a never-ending source of mystery to me how orgs that see security in terms of selecting which commodity products to buy expect to defend against advanced attackers that show up with custom tooling and exploits. If the product isn't export-controlled, how good is it really?

3 11 29 5K 6

Zion Leonahenahe Basque @mahal0z

7 months ago

It's finally here! Ghidra 11.3 dropped with built-in support for Python 3 through Pyhidra. Let's go! github.com/NationalSecuri…

11 133 639 35K 188

LaurieWired @lauriewired

7 months ago

Ghidra 11.3 is OUT! PyGhidra is the new feature to be excited about. It’s a Python library providing direct access to the Ghidra API. I expect this to massively increase Reverse Engineering tool development, as it significantly reduces the barrier to entry.

13 121 892 48K 210

Download Gif

Dino A. Dai Zovi @dinodaizovi

8 months ago

This right here: "The USG can and should respond to China and other adversaries using all instruments of power within its reach. But China isn’t the only adversary capable of compromising these targets, so without fixing the underlying vulnerabilities another attack is just a…

chrisrohlf @chrisrohlf

8 months ago

1 0 2 3K 0

1 4 5 2K 0

Wyatt @whyitfor

8 months ago

Finance ppl expect more M&A than IPO in #cybersecurity in 2025: “Midmarket cyber vendors are ripe for consolidation” Increased CISO spending is expected to reinforce this trend, as CISOs are looking for fewer vendors with more tools wsj.com/articles/cyber…

0 0 0 26 0

Julien Vanegue @jvanegue

8 months ago

List of 340 CVEs detected to be exploited in the wild in the past many years. Thanks to whoever compiled that! This is a gold mine for vulnerability research.

j j @mistymntncop

8 months ago

List of 340 CVEs detected to be exploited in the wild in the past many years. Thanks to whoever compiled that! This is a gold mine for vulnerability research.

0 2 13 1K 4

0 1 0 628 4

lukas seidel @pr0me

8 months ago

happy new year everyone 🎉 to continue the tradition, here is some of my favorite firmware & embedded security research of 2024: Defeating the new Raspberry Pi's RP2350 Security Features [1] Reversing and Hacking Firmware of an in-orbit Satellite to Re-establish Lost…

1 20 64 7K 39

Download Image

Dino A. Dai Zovi @dinodaizovi

10 months ago

We used this exact phrase at @Square in 2014 when I first started. We took on the as much of the complexity around PCI compliance as possible for Square sellers so that they didn't have to buy security products for their business. We were the secure product for their business.

Phil Venables @philvenables

10 months ago

0 14 48 9K 19

Download Image

0 1 10 2K 2

Wyatt @whyitfor

11 months ago

This speaks to why product certification alone will never be enough. Many certs are granted without actually validating the correctness of the implementation. Adding a mitigation != correctly implementing it

Mick Douglas 🇺🇦🌻 @bettersafetynet

11 months ago

4 3 49 8K 3

0 0 1 23 0

Mick Douglas 🇺🇦🌻 @bettersafetynet

11 months ago

Nothing gets me angrier when someone misrepresents a technology. Part of why I pen test is to prove things. What sort of things you ask? Vendor claimed they did Mutual TLS Turns out... they had a 'unique' definition of what this is. 1