Blockchain Security Brief
The weekly record of web3 darkest hours

Tuesday, December 16, 2025

Top Exploits

Why would attackers evolve when DeFi keeps doing the hard work for them?

  • USPD didn’t get hacked - they got pre-owned. Seventy-eight days before a single dollar was stolen, a stranger quietly claimed the admin slot during deployment, installed a shadow proxy, spoofed every on-chain indicator, forwarded every call to the real code, and waited. The audits were clean. The math was correct. The proxy was possessed. When the attacker finally woke up, they mint-looped 98 million USPD out of 3,121 ETH, drained 232 stETH, dumped $300K through Curve, and left $1M sitting untouched on-chain like a trophy. This wasn’t an exploit - it was a hostile protocol takeover executed 24 seconds after deployment. July’s CPIMP war room saved dozens of protocols, but USPD launched two months later still vulnerable. (Read more)

  • Prediction markets wanted to be Vegas. Instead, they recreated the skim. Six days after Paradigm led a $1B round into Kalshi, their in-house researcher published findings that cut Polymarket’s most important metric - lifetime volume - from $25B to $13.5B overnight. The “bug” wasn’t wash trading or fraud; dashboards had been double-counting OrderFilled events for years, and everyone in the data chats knew it. The timing? Impeccable. Polymarket was pitching a $12-15B raise; Kalshi had just doubled to $11B; CNN and CNBC were integrating Kalshi odds into live broadcasts. What used to require undercover agents, casino raids, and accounting stings now takes one blog post and a transaction simulator. (Read more)

Rekt Audit Broker

The big city sleeps, but the code never does. Neither do the crooks.

Rekt Audit Broker connects protocols with top-tier security firms. One request, multiple bids. Faster audits, fairer pricing, trusted partners.

Audits are cheaper than funerals.

Deep Dives

Securing Blockchain in Banking: TradFi’s Journey from Bitcoin to DeFi (7 min read)
Bitcoin was born as a revolt against the banks, but now the banks are strip-mining its tech to fix their own broken plumbing. Cross-border payments, identity rails, and KYC-driven “decentralized” credentials have become TradFi’s gateway drugs into blockchain - fast, cheap settlement wrapped in layers of surveillance-grade identity. What started as an escape hatch from centralized finance is slowly being rebuilt into its next infrastructure layer, where blockchains handle the transfers and institutions handle the keys, the rules, and the risks.

Zero Trust IAM for Web3 Protocols (8 min read)
Crypto infra still gets wrecked the same way Web2 did: stale keys, ghost admin accounts, and “temporary” privileges that last until the next post-mortem. Zero-trust IAM flips the model - every request verified, every role scoped to reality, every signer accountable - across cloud consoles, CI pipelines, multisigs, guardians, and upgrade keys. In a world where your protocol’s fate hinges on whoever still has a seed phrase in a browser extension, identity isn’t convenience. It’s the real attack surface, and the only line between institutional trust and instant liquidation.

What is Fuzzing in Web3 Security? A High Level Overview (9 min read)
Manual audits catch what humans can imagine; fuzzers catch everything else. By mutating calls, twisting state transitions, and slamming a protocol with thousands of deranged input sequences, fuzzing exposes the financial edge cases, invariant leaks, and multi-step failure paths that no developer - and no audit team - ever wrote a test for. In DeFi, where one broken assumption can turn a curve into a drain, fuzzing has quietly become the closest thing we have to a machine-generated attacker: relentless, unbiased, and disturbingly good at finding the places your system snaps under pressure.

Smart Contract Incident Report: Legacy Bridge Vulnerability (5 min read)
A forgotten, un-bricked 2024 Bridge proxy stayed live on-chain after April’s vulnerability fix, letting an attacker harvest unlimited ERC20 approvals from one Ethereum wallet that had bypassed standard allowance caps. No native assets were touched, and all legacy proxies are now forcibly upgraded to a null implementation. The breach came down to one miss: a legacy contract left breathing when it should’ve been buried.

Catching malicious package releases using a transparency log (7 min read)
Sigstore’s upgraded rekor-monitor turns transparency logs into early-warning systems, alerting maintainers the moment a hijacked identity is used to ship a malicious package. With v2 log support, certificate checks, TUF-backed key retrieval, and a plug-and-play GitHub workflow, it makes supply-chain compromise impossible to hide - and finally gives developers a siren when their own signing keys betray them.

The End of Free Trust


We built the internet on a comforting illusion - that trust is free.
Free to issue.
Free to verify.
Free to depend on.

Credentials, licences, diplomas, onboarding checks - all treated like harmless paperwork drifting through digital pipes. Meanwhile every organisation quietly burns time and money running the same verifications over and over, pretending the system magically balances itself.

But trust was never free.

The costs were just buried under the floorboards.

Issuers carry the compliance load without any guarantee of compensation.
Verifiers repeat the same checks a hundred others already performed.
Holders endure endless identity loops like characters trapped in a Kafka side quest.

The problem isn’t fraud.
It’s that trust has no economic memory - no model that aligns who creates it, who depends on it, and who pays for sustaining it.

And now that generative AI can forge “evidence” faster than institutions can review it, verification has become the new scarcity.
Information used to be abundant and reliable; now it’s abundant and untrustworthy.
The only thing with value anymore is proof.

Here’s the shift no one can ignore: trust is becoming a programmable asset, not a free assumption.

Some ecosystems have started treating it that way. One example is cheqd’s work on credential payments - not as a product pitch, but as a signal that trust finally has a price tag tied to its utility.

Issuers get rewarded for maintaining accurate credentials. Verifiers pay only when they derive value. Holders reuse credentials instead of obtaining them repeatedly like digital passports with infinite renewal fees.

When trust becomes programmable, everything changes.
Costs align with benefits.
Value flows where the work actually happens.
And the system stops pretending verification is an act of charity.

What’s even more interesting is how quickly others are building on top of this shift. Several platforms are already using cheqd’s payment primitives to anchor incentives directly into credential flows, turning trust into something measurable instead of a vague handshake between institutions. 

The truth is simple: the internet we’ve been using runs on blind faith and duplicated effort.

And that model collapses the moment AI turns misinformation into a commodity.

The future won’t ask “Can we trust this?”

It’ll ask “Who paid for the trust that makes this provably real?”

And for once, the system will actually have an answer.

*Sponsored article

Other Security Stories

Portugal Clarifies Cybercrime Law. A new safe-harbor provision now protects good-faith security research under strict conditions - requiring no harm, no profit, rapid disclosure, and tight limits on testing methods.

Anonymous Phone Service Launches in the US. Phreeli now offers mobile numbers with only a ZIP code and zero-knowledge proof billing, letting users pay without exposing identity while sidestepping the data-harvesting practices of mainstream phone providers.

NGate Malware Skims NFC Cards Through Android Phones. A new Android trojan tricks users into “verifying” their bank cards, skimming NFC data and PINs via fake apps and phishing lures before criminals drain accounts through ATM withdrawals, now spreading beyond Czech banks into Poland.

North Korean APT Dev Rig Exposed by Its Own Malware. A LummaC2 infection exposed a DPRK malware developer’s machine, linking stolen credentials and phishing infrastructure directly to the $1.4B ByBit heist - an accidental leak revealing how state hackers build, launder, and deploy their tools.

React2Shell Drives Widespread Miner and Malware Deployment. Threat actors are abusing the RSC bug at scale to drop XMRig miners and new Linux implants across multiple industries, with automated tooling flooding servers as exploitation surges globally.

New Tools and Projects

TestMachine Predator: TestMachine raised $6.5M to expand Predator, its AI-driven blockchain security platform for automated vulnerability scanning, on-chain risk analytics, and exploit detection across DeFi protocols.

qLABS Quantum-Sig Wallet: qLABS and Quantum Inc. unveiled Quantum-Sig, a post-quantum cryptography-enabled smart contract wallet designed to protect digital assets (ETH, SOL, USDT/USDC, etc.) with quantum-resilient signature mechanisms to defend against future quantum attacks.

HAI Group CORE.3 Platform: HAI Group launched CORE.3, a Web3 risk intelligence and loss-metric platform introducing probability-of-loss metrics and deeper probabilistic risk scoring to help security teams quantify and monitor protocol exposure.

ConneX Cross-Chain Security Resolver: ConneX is a cross-chain security analysis tool that automatically resolves opaque bridge transaction pairs using semantic and LLM-assisted pruning, significantly improving traceability and vulnerability detection across multi-chain bridges.

SecureSign SDK (EIP-6963 sandboxing): A new mobile Web3 security framework that uses emulated provider sandboxing for EIP-6963 to isolate dApp interactions and mitigate click-jacking, overlay, and skimming attacks - offering improved transaction integrity and UX for mobile users.

Rekt Flashback

One year ago, GemPad’s “secure” lockboxes snapped open like cheap luggage when a missing reentrancy guard let an attacker siphon $1.9M across Ethereum, BNB Chain, and Base. One malicious token, one reentrant callback, and suddenly fees became free LP locks, and LP locks became a one-way liquidity slide. BPay, Munch, Nutcoin and half a dozen others watched their “locked” assets drain while GemPad’s no-code, pre-audited templates turned out to be security theater on a ticking time bomb. By the time analysts traced the flaw, the attacker’s funds were already dissolving into mixers, leaving projects to explain how their locks weren’t locks at all.

Memes and Videos

North Korea’s Most Destructive Hack: Dark Seoul

South Korea woke up to find its hard drives erased and its enemy inside the walls months earlier. Dark Soul wasn’t a hack - it was a silent occupation. One RAT, one wiper, and a country on its knees.

Source: Cybernews

Source: lynk0x

Want to partner with us?

Skip the bots, hit the brains.

Get your message in front of the sharpest, most battle-tested crowd in crypto.

If they notice you, the whole space will. [Partner with us]

We provide an anonymous platform for whistleblowers and DeFi detectives to present their information to the community. All authors remain anonymous. 
We are all rekt.

Keep Reading

No posts found