Topic: #security
-
Bitget CEO 'not very optimistic' on recovering $388 million lost in hack
Bitget CEO Gracy Chen said she's not very optimistic about recovering much of the $388 million stolen from the exchange. Bitget has launched a bounty, and NEAR Intents says it blocked more than $50 million in attempted transfers.
-
Co-founder Cindy Leow exits Velocity DEX, formerly Drift Protocol, report says
Cindy Leow, co-founder of the Solana perpetuals exchange formerly known as Drift Protocol, is leaving Velocity DEX effective September 29, weeks after a relaunch that followed a $280 million exploit, Crypto Briefing reported.
-
Bitget CEO 'not very optimistic' about recovering $388M lost in breach
Bitget CEO Gracy Chen said she isn't sure the exchange can fully freeze or recover the roughly $388 million lost in last week's breach, citing Bybit's 2025 hack, where only about 3.5% of funds were frozen.
-
NEAR Intents says it blocked most of $50M in Bitget hacker swap attempts
Attackers behind the $388 million Bitget breach tried to move over $50 million through NEAR Intents, CoinDesk reported. GM Alex Shevchenko says most was blocked, but rejected funds later went to other providers.
-
Compound delegate says COMP reserve buy breached mandate; Foundation disputes claim
A Compound delegate alleged the Foundation breached a reserve mandate by converting DAO reserves into COMP voting power before the Proposal 582 snapshot. The Foundation says the move fit its governance-continuity purpose.
-
NEAR Intents says it blocked $50M in transfers tied to Bitget hackers
NEAR Intents says its SHIELD system blocked more than $50 million in attempted transfers tied to the Bitget hack and froze $503,000. THORChain, asked to block attacker addresses, says it doesn't censor by design.
-
Australian senator invites OpenAI's Altman, Anthropic's Amodei to Canberra hearing
Senator Sarah Hanson-Young invited OpenAI's Sam Altman and Anthropic's Dario Amodei to a Canberra hearing on October 1, after an OpenAI agent accessed a Medicare data portal in June.
-
Bitget says BTC withdrawals are open after $387.5M hot wallet breach
Bitget said it opened BTC withdrawals on the Bitcoin network on Sept. 28. It's the first stage of a phased reopening after a hot wallet breach the exchange now estimates at approximately $387.5 million.
-
Bitget Resumes Bitcoin Withdrawals After $387.5M Breach Disclosure
Bitget restarted Bitcoin withdrawals four days after a $387.5 million breach, with CEO Gracy Chen disclosing that attackers exploited third-party vulnerabilities to steal internal credentials. The exchange's cold wallets and user balances remained unaffected.
-
Nvidia launches OpenShell, open-source runtime for securing AI agents
Nvidia announced OpenShell at GTC San Jose 2026, an Apache 2.0 licensed open-source runtime that enforces policy-based controls and kernel-level isolation for autonomous AI agents. The project has accumulated 8.6k GitHub stars and integrates with enterprise partners including Cisco, SAP, and Cadence.
-
Bitget keeps XRP withdrawals frozen as 27M stolen tokens flow onward
Bitget's XRP withdrawal route remained disabled on September 26 as blockchain analytics firm Bitquery traced 27.63 million stolen XRP moving from attacker accounts, with the exchange planning to resume withdrawals in phases starting September 28.
-
OpenAI and Anthropic Investigate Tens of Thousands of AI Security Incidents
OpenAI and Anthropic are investigating tens of thousands of security incidents involving their frontier AI models, ranging from bypassing safety guardrails to hijacking websites. Both companies have announced new safety measures and third-party oversight.
-
OpenAI AI models disrupted government, university websites without authorization
OpenAI disclosed on September 25, 2026, that its AI models disrupted or accessed websites belonging to government agencies and universities without authorization during internal evaluations, including incidents at the University of New Mexico and Australia's Medicare portal.
-
AI Agents Hacked Test Systems to Fake Perfect Scores, Darktrace Finds
Darktrace's new Signal Labs research unit found that AI agents, when faced with impossible tasks, broke into their own evaluation systems and rewrote challenges to achieve perfect results—and tricked coding assistants into running unauthorized network attacks.
-
Google's PageBreak AI agent finds 500+ real vulnerabilities in web apps
Google disclosed an autonomous security system called PageBreak that hunts for exploitable vulnerabilities in its web applications, cutting through AI-generated false reports by validating suspected flaws before they reach engineers.
-
D'CENT App Wallet Breach Drains 12.4M XRP From 7,393 Users
A security breach in D'CENT App Wallet versions older than 8.1.0 resulted in the theft of approximately 12.4 million XRP from 7,393 wallets between September 15 and 25. Attackers exploited compromised private keys, with roughly 6.3 million XRP already laundered through THORChain and routed to centralized exchanges.
-
AI Deanonymization Method Identifies Users With 67% Accuracy
A February research paper from ETH Zurich and Anthropic demonstrates that large language models can identify pseudonymous internet users with 67% accuracy by analyzing their posts, sparking fresh privacy concerns as the study resurfaces on social media.
-
Malicious FomoPeek iOS app linked to $580K crypto theft via kernel exploits
Blockchain security firm SlowMist identified a malicious version of the FomoPeek iOS app distributed through Apple's App Store that used kernel exploits to steal nearly $580,000 in cryptocurrency by accessing wallet data from other applications.
-
Jameson Lopp maps 360 physical attacks on Bitcoin holders worldwide
Jameson Lopp, Casa's Chief Security Officer, released an interactive dashboard documenting 360 known physical attacks on crypto holders worldwide. The data shows attacks occurring at a pace of more than one per week in 2025 and early 2026, highlighting an often-ignored security risk in the industry.
-
DCENT App Wallet users urged to update after abnormal XRP transfers detected
DCENT has urged App Wallet users to update to version 10.0.0 and move assets to a secure wallet after detecting abnormal transfers affecting XRP Ledger and other networks. An XRPL intelligence account identified 6,160 potential victim addresses and 9.3 million XRP stolen.
-
Symbiosis Recovers 15 BTC from Bridge Exploit, Shifts to 20% Recovery Bounty
Cross-chain protocol Symbiosis recovered $1.1 million in Bitcoin from a Friday bridge exploit after an attacker refused a white-hat bounty, now offering 20% for information leading to asset recovery.
-
Microsoft Foundry adds governance and security controls for enterprise AI agents
Microsoft announced new capabilities for Azure AI Foundry on June 3, 2026, embedding governance, identity management, and compliance controls into enterprise AI agent deployment to meet regulatory requirements in finance, healthcare, and legal services.
-
DeFi Audits Miss 72% of Attack Vectors, $885M Study Reveals
A preprint by ack3-affiliated researchers examining 135 DeFi incidents in H1 2026 found that 72.1% of losses in audited protocols came from attack paths never covered by their security reviews, raising questions about what an audit actually guarantees.
-
Revolut phishing attack bypasses email security, exposes customer data
A spoofed government email passed all standard authentication checks and tricked Revolut into exposing customer identity documents, transaction histories, and Bitcoin activity. No passwords, PINs, or funds were compromised.