The $110M Balancer hack is a tragedy, and this widely-shared image showing "10+ audits" is leading many to the wrong conclusion: that audits are useless. This isn't an audit crisis; it's a data transparency crisis. This static list is misleading because it hides all the critical context. It's a "stamp" of security that fails to answer the most basic questions a user or developer should ask: 1. Is it the right code? Was the deployed code that was hacked the exact same code that was audited? Is there a verifiable link, like a commit hash, to prove it? 2. What did the reports say? Were there unresolved high-severity issues that the team acknowledged but didn't fix? There's no visibility on the actual content of the audits. 3. What happened after the audit? Was the code upgraded? Did a change get pushed that invalidated the original report? The takeaway isn't that "audits are broken." The takeaway is that we're relying on a broken, static system to prove them. We don't need more lists; we need a live, verifiable data layer that provides the real context behind the "audited by X" claim.

A quieter quarter does not mean a safer one. Exploit wallets are still active, and laundering activity on Tornado Cash has tripled since June. The real signal of maturity will come when losses fall because defenses improved, not because attackers paused. And @TrustblockHQ is committed to building a safer ecosystem, one verified block at a time.

Centralized exchanges were hit hardest, losing $182M in Q3. DeFi projects followed with $86M in losses. Most incidents involved phishing or social engineering targeting multisig and hot-wallet access. The balance has shifted from technical exploits to credential and workflow compromise, a sign that defenses must expand beyond code security.

Two incidents shaped the month: ■ @UXLINKofficial lost roughly $44M after an attacker minted 10 trillion unauthorized tokens. ■ @swissborg suffered $41.5M in losses through a breached API provider. Smaller exploits included a $13.5M phishing theft on @VenusProtocol and about $10M combined from @yalaorg and @Griffin_AI. No single event crossed $100M, but the clustering of mid-sized breaches kept risk levels high.

September marked a paradox. It delivered a record number of million-dollar exploits while overall losses declined. 16 hacks crossed the $1M mark, more than any month on record. Total September losses reached $127M, spread across about 20 incidents. The scale is smaller, but the frequency is higher, showing that risk remains distributed across the ecosystem.

Losses from code vulnerabilities plunged from $272M in Q2 to just $78M in Q3. Phishing incidents caused less damage even though frequency remained constant. Attackers are adapting. They are shifting focus from contract logic to wallets, exchanges, and human access points. The weak link is moving closer to users and operations.

2025 has been one of crypto’s most volatile years for security. But for the first time in a while, there is progress. Losses from hacks and exploits fell 37% in Q3, from $803M to $509M, according to @CertiK. That is a 70% drop compared with Q1. The catch: September still recorded the highest number of million-dollar hacks ever. Here's what the last quarter revealed about the state of crypto security 🧵

The ecosystem grows naturally when users can see: ■ How a protocol is secured ■ How ownership is distributed ■ How interactions are verified Each of our partners contributes a piece to this new standard. Explore security insights at a glance: trustblock.run

At the end of the day, prioritizing security can still be a fun endeavour. @0xBilel sums up the discussion by illustrating how verifiable proof can turn into lasting confidence.

But code safety is not the whole picture. Ownership behavior changes outcomes. @cecilialsn sheds light on why @bubblemaps makes wallet clusters and bundling visible.

Protection should feel like part of the flow. @mrccss talks about how @mywebacy adds simulation and risk checks at the moment of intent.

Surface logs miss storage-level risk and upgrade conflicts. @bilaldanishm delves into the process of extracting deep storage data to power accessibility at @SmartMuv.

Explorers are shifting from raw data views to context-rich safety layers. @mike_krupin explains how @blockscout is helping users judge risk at a glance across EVM-based blockchains.

The question we posed to the ecosystem was simple: can Web3 security work if users can’t see it? Across a live AMA and a week-long puzzle campaign with @blockscout, @bubblemaps, @mywebacy, and @SmartMuv, one answer kept coming up. Here’s what each team had to say 🧵

We're super excited to be joining the @TrustblockHQ auditor network! We believe that security should be transparent and verifiable, and now it is. app.trustblock.run/auditor/q…

We’re proud to be part of Trustblock! 🤝 Trustblock is the audit infrastructure layer for Web3 - indexing thousands of verified audit reports from top firms into real-time, composable data layers.

The $2-billion loss recorded in 2025 reflects both the scale of DPRK operations and the gaps in crypto security. Improvement depends on disciplined practices: ■ MPC wallets and multi-person approvals ■ Continuous audits and time-locked governance ■ Regular phishing awareness for teams ■ Real-time on-chain monitoring Security must operate continuously rather than in single snapshots. That’s the principle we build on at @Trustblock_HQ.