We help secure the world’s most targeted organizations and products. We combine security research with an attacker mentality to reduce risk and fortify code.

New York, NY
We launched Patch the Planet with OpenAI, factored hundreds of weak RSA keys with a new polynomial technique, and bypassed every AI skill scanner we tested. Plus 10 new public reviews, gosentry, and more. June Tribune: mailchi.mp/trailofbits/june-…
1
13
77
6,933
We built local backdoors for @signalapp, @1Password, @SlackHQ, and @googlechrome using a “data” file their integrity checks ignored.
10
107
990
79,840
Today, we are disclosing LeftoverLocals, a vulnerability that allows listening to LLM responses through leaked GPU local memory created by another process on Apple, Qualcomm, AMD, and Imagination GPUs (CVE-2023-4969) buff.ly/48RDP68
11
214
915
306,698
Today we’re releasing weAudit, the VSCode extension we use during secure code reviews to collaboratively take notes and highlight code regions. blog.trailofbits.com/2024/03…
17
149
558
66,064
For the last year, a 9-person team from @trailofbits has deeply studied the security of blockchains for @DARPA. Today, our analysis and tools are public: blog.trailofbits.com/2022/06…
50
261
564
Today, we are releasing RPC Investigator, made for exploring RPC clients and servers on Windows. This .NET application builds on the NtApiDotNet platform, adding features that offer a new way to explore RPC blog.trailofbits.com/2023/01…
5
157
414
59,950
Your code might be vulnerable! Our cryptography team has discovered a number of Fiat-Shamir vulnerabilities affecting proof systems such as Bulletproofs and PlonK. Check out this blog series for details and contact us if you think your codebase might be… blog.trailofbits.com/2022/04…
12
123
372
We’re thrilled to announce our new Testing Handbook, which gathers insights we gained over years of experience using static and dynamic analysis tools. It goes beyond standard documentation, focusing on giving the right answers rather than all the answers. blog.trailofbits.com/2023/07…
1
80
340
51,506
Microsoft didn’t sandbox Windows Defender, so I did blog.trailofbits.com/2017/08…
5
206
289
It's easy to find bugs when you know how to build the right tools. Check out our blog to learn how to model vulnerabilities with Binary Ninja's MLIL and SSA form. blog.trailofbits.com/2018/04…
2
119
290
Check out our _accessible_ Meltdown and Spectre explainer, made for developers without a background in computer architecture. No awkward analogies, we stick to the real details. blog.trailofbits.com/2018/01…
156
280
How to instantly exploit 2000 binaries with Binary Ninja blog.trailofbits.com/2016/06…
1
188
280
We’re releasing a Ghidra extension, BTIGhidra, that automatically recovers types with inter-procedural analysis and enhances decompilation for improved reverse engineering comprehension. buff.ly/3uheYdl
3
66
274
20,429
We hacked Gemini CLI, Vertex AI, Assistant, and other AI systems by embedding prompts into images that are not visible to users.
8
56
277
57,064
We've just released secure-contracts.com, a revamp of our guide to building secure contracts! It contains security guidelines, best practices, tool tutorials, and many other resources.
6
67
257
25,945
Over the years, we have accumulated advanced knowledge and guidance for writing better smart contracts. We are sharing this knowledge in the first release of building-secure-contracts: github.com/crytic/building-s… Check out the repo to learn about best practices and tooling!
6
76
245
The $1.5B Bybit hack marks a new era in cryptocurrency security. Attackers have moved beyond technical exploits to sophisticated operational attacks. Read our initial analysis of this historic breach and its industry-wide implications: blog.trailofbits.com/2025/02…
9
67
252
330,066
The Linux kernel 6.10 introduces the mseal syscall for memory protection. Discover its unique features, how it differs from prior schemes, its kernel implementation, and the userspace exploits it prevents. hubs.la/Q02VNW-30
2
56
251
43,720
Shoutout to @1Password, @signalapp, and @SlackHQ for patching this vulnerability before our public disclosure, demonstrating how responsible disclosure makes the entire ecosystem safer.
5
11
244
6,848
We audited the @golang standard cryptographic library, used by thousands of libraries and millions of users. Here's what we found and some key takeaways 🧵
6
40
249
24,759
We won second place in @DARPA's AIxCC with Buttercup! Our CRS is now open source for everyone to use.
7
44
248
18,053
Today we released Echidna, our next generation EVM smart fuzzer at #EthCC. It's the first-ever fuzzer to target smart contracts! blog.trailofbits.com/2018/03…
6
133
232
Destroying x86_64 instruction decoders with differential fuzzing blog.trailofbits.com/2019/10…
4
98
217
Since the Balancer hack on Monday, we've been working around the clock to help the Balancer team understand the bug and its implications. We've verified the bug's cause and have independently verified that the exploit does not work on Balancer V3. Once the dust settles, we plan on doing a retrospective and sharing more information including long-term, strategic guidance on how to avoid similar bugs.
6
12
215
44,231
As smart contract security evolves, property-based fuzzing has become a go-to technique for developers and security engineers. To help the community define properties, we are releasing a set of 168 pre-built properties that can be used to guide Echidna. blog.trailofbits.com/2023/02…
5
45
189
38,619
Most border crossings depend on cryptographic protocols most people have never heard of. You should know the technology and security risks behind your passport.
4
18
188
25,251
We’re launching a new service: invariant development. We’ll identify, implement, and test security-critical invariants to prevent bugs & secure your codebase over the long term. Plus, we’ll upskill your team to write their own invariants! buff.ly/3PIbjM7
6
38
180
51,593
Reverse engineer 200 binaries with the mechanical efficiency of symbolic execution blog.trailofbits.com/2017/05…
88
176
We've discovered yet another MCP attack technique! Attackers can hide malicious payloads using ANSI terminal escape codes. When your AI agent processes these invisible instructions, it can leak data or compromise your supply chain without you seeing anything suspicious.
6
49
185
25,824
We found critical flaws in common TSS libraries, a cryptographic protocol for distributed key generation and signing. We’ve released ZKDocs.com to help secure the rapidly advancing field of ZKPs, TSS, and similar schemes. blog.trailofbits.com/2021/12…
6
56
179
Warning: @lfgexchange is falsely claiming to have worked with us on an audit. The report on their page is fake. If you want to verify the authenticity of a @trailofbits report, find it on our publications repo, the authoritative source straight from us. github.com/trailofbits/publi…
8
37
169
87,067
MCP Vulnerability: AI systems can be manipulated to silently leak your entire conversation history to malicious servers.
5
40
173
22,368
Passkeys are the most important security technology of the past 10 years. You should know how they work. 🧵
12
41
175
19,878
There's a dark side to compiler optimizations that can inadvertently cause information leaks or remove code critical to security. Read about the research being done to mitigate this risk and reduce the exposure to code-reuse attacks in software. blog.trailofbits.com/2022/03…
3
47
147
Devirtualizing C++ with Binary Ninja blog.trailofbits.com/2017/02…
102
152
Protecting Software Against Exploitation with DARPA’s CFAR blog.trailofbits.com/2018/09…
1
76
149
We published a technical summary of the "AMD Flaws" so they can be of use to the security community without the distraction of the surrounding disclosure issues. blog.trailofbits.com/2018/03…
3
122
142
The @raft_fi issue is complex and we're actively investigating it. We've offered to help their team however we can. Here's some of what we know so far:
2
21
152
51,108
Fuzz smarter, not harder, with our newest Testing Handbook chapter that will take your fuzzing skills from 0 to 60. buff.ly/3w8PAa1
1
32
146
14,655
Today's release of our Echidna smart contract fuzzer is more than 20x faster, thanks to multicore support and optimized coverage collection. github.com/crytic/echidna/re…
4
33
144
22,841
If you missed our @BurpSuite webinar ft @albinowax, you can watch the recording now on Youtube. buff.ly/4ds1dd7 We cover: Advanced web research techniques using Burp Suite How to discover ideas and targets Optimize your setup & utilize Burp tools in various scenarios ++ Q&A with @albinowax
1
40
148
19,053
Is your centralized exchange, bridge, or L2 client using block delays to determine transaction finality? If so, it may be vulnerable to re-orgs, double-spend attacks, and stolen funds. Our new guide to blockchain finality helps you avoid these attacks. blog.trailofbits.com/2023/08…
4
28
137
15,841
Before its prod launch, we used every crowbar in our toolbox to break into @WhatsApp’s Private Processing TEEs, and starting from a compromised hypervisor gave plenty of leverage. The enclaves yielded to injected ACPI tables, environment variables and evil file names; but now the system is stronger for it.
3
14
137
40,056
How Threat Modeling Could Have Prevented the $1.5B Bybit Hack. Our blog explores one of our most popular but rarely published report types, and how adding threat modeling to your organization can save you from becoming the next billion-dollar headline. blog.trailofbits.com/2025/02…
2
28
124
13,438
As a summer intern, Jason An upgraded Pwndbg with Binary Ninja integration and Go structure dumping, bringing an IDE-like experience to stripped binary analysis. buff.ly/3Xvcgwy
1
24
130
12,337
Buttercup won the $3M second prize at DARPA's AIxCC. We found 28 vulnerabilities across 20 CWEs with 90% accuracy at just $181/point, achieving this with exclusively non-reasoning LLMs.
2
15
133
11,581
Today we're releasing Attacknet, a new tool in the blockchain security arsenal. Built in collaboration with the @Ethereum Foundation, it uses Chaos Engineering to test the most challenging network conditions imaginable for fault tolerance blog.trailofbits.com/2024/03…
2
30
132
17,577
We are now accepting applications for our annual summer internship in the following areas: • AI/ML • Application security • Blockchain • Cryptography • Operations Learn more in 🧵
4
13
131
14,224
We have a working proof-of-concept exploit for ‘Whose Curve is it Anyway?’ — NSA’s bug in Microsoft’s Crypto API. Read on for our explainer: blog.trailofbits.com/2020/01…
2
59
124
Upgrading smart contracts can introduce new bugs, risking millions of dollars. We've developed Diffusc, a differential fuzzer that compares two smart contracts to uncover unexpected differences in behavior before an upgrade is deployed. buff.ly/3rq00zW
4
37
122
19,821
Earlier this week, @UncipheredLLC disclosed that BitcoinJS, the most widely used JavaScript library for bitcoin wallets, relied on weak randomness until 2014. This issue puts millions of wallets at risk. Here’s what we know:
Experts discover flaw leaving $1 billion in bitcoin and other cryptocurrencies exposed for stealing from early software wallets. Free link to my story in the The Post: wapo.st/478Av5W #bitcoin #doge #infosec
5
28
123
61,009
Async-unsafe signal handlers are at the core of the recent regreSSHion vulnerability (CVE-2024-6387). We published a new CodeQL query that searches for this often overlooked class of bugs. github.com/trailofbits/codeq…
1
29
123
11,804
If you're fuzzing C/C++ code and need more customizability, our new Testing Handbook chapter shows you exactly how to set up and use LibAFL as a libFuzzer drop-in and a Rust library. (More in 🧵)
1
16
118
13,539
We’ve built many high-impact tools that we use for security reviews. But mastering them can take time. So we're bringing the mastery to you: we're going to be livestreaming tool workshops on our Twitch and YouTube channels! blog.trailofbits.com/2022/11…
6
29
115
Clang isn't a toolsmith's compiler. PASTA tries to fix this by providing safe-to-use C++ and Python wrappers to the Clang AST. PASTA also answers questions that Clang can't, like how parsed tokens relate back to macro expansions and files. Learn more: blog.trailofbits.com/2023/07…
1
32
101
23,461
Magnifier is a UI that helps reverse engineers explore decompiled programs interactively without all the manual note-taking. Read about this excellent work from our intern, @tcode2k16! blog.trailofbits.com/2022/08…
1
26
110
Trail of Bits donates $100,000 to support young researchers through @SummerC0n blog.trailofbits.com/2018/06…
3
30
108
Prompt injection dominates AI security discussions, but little public research exists on writing powerful, discreet, and reliable exploits until now 🧵
1
19
109
16,076
Reverse Taint Analysis Using Binary Ninja blog.trailofbits.com/2019/08…
58
103
We are now accepting applicants for our summer internship program! We will be hiring approximately 10-15 interns across our research, engineering, and assurance practices boards.greenhouse.io/trailof…
5
24
94
43,661
Today we're releasing Caracal, our new static analysis tool for Starknet smart contracts. It has 10 detectors that detect reentrancies and other vulnerabilities, two printers, and more! github.com/crytic/caracal
2
25
99
10,463
Breaking Down Binary Ninja’s Low Level IL blog.trailofbits.com/2017/01…
63
98
An accessible overview of Meltdown and Spectre, Part 2 blog.trailofbits.com/2018/03…
1
54
97
Rewriting Functions in Compiled Binaries blog.trailofbits.com/2019/09…
43
99
Let’s talk about CFI: Microsoft Edition blog.trailofbits.com/2016/12…
2
78
95
DARPA's AIxCC finals: 7 autonomous AI systems are competing RIGHT NOW to find and patch vulnerabilities in critical open-source programs like the Linux kernel, SQLite, and cURL. 🧵
3
24
97
16,828