🚀 Anthropic's Report Highlights Vulnerabilities in Smart Contracts

According to Foresight News, Anthropic has released a report detailing the performance of its models, Claude Opus 4.5, Claude Sonnet 4.5, and GPT-5, on the SCONE-bench benchmark. This benchmark includes 405 real-world attacked contracts from 2020 to 2025. The models identified exploitable vulnerabilities worth approximately $4.6 million in contracts attacked after the knowledge update time of March 2025.

In addition, a simulation test was conducted on 2,849 recently deployed contracts with no known vulnerabilities. During this test, Sonnet 4.5 and GPT-5 each discovered two new zero-day vulnerabilities, potentially leading to a combined loss of $3,694. Notably, the API cost for GPT-5 was $3,476.

#Anthropic #ClaudeOpus #ClaudeSonnet #GPT5 #SmartContracts #Vulnerabilities #SCONEbench #ForesightNews #ZeroDay #ExploitableVulnerabilities #BlockchainSecurity

🚀 AI Agents Demonstrate Significant On-Chain Attack Capabilities

According to BlockBeats, recent research by Anthropic reveals that AI agents possess substantial on-chain attack capabilities. In simulated tests on smart contracts that were hacked between 2020 and 2025, Claude Opus 4.5, Sonnet 4.5, and GPT-5 collectively replicated vulnerabilities valued at approximately $4.6 million. Additionally, while scanning 2,849 contracts with no known vulnerabilities, the models discovered two new zero-day vulnerabilities and successfully simulated profits.

The study highlights that the profitability of AI-driven on-chain attacks has doubled approximately every 1.3 months over the past year, indicating that AI technology is fully capable of autonomously exploiting vulnerabilities for profit.

#AIagents #onchainattacks #smartcontracts #vulnerabilities #ClaudeOpus #Sonnet #GPT5 #zeroDayVulnerabilities #AIdriven #cybersecurity #blockchain #profits #AIresearch

🚀 Security Flaws Detected in Two Major Trading Platforms

According to BlockBeats, SlowMist's Cosine reported on the X platform that the SlowMist team has identified serious vulnerabilities in two trading platforms, which directly impact the security of funds. Attempts to contact the platforms have been unsuccessful, and public outreach has also yielded no response. Cosine noted that one platform has a 24-hour trading volume of $3.7 billion, while the other has a volume of $240 million. A tweet from the SlowMist team account indicated that they had proactively reached out to a trading platform named 'ICRYPEX Global' the previous day.

#SecurityFlaws #TradingPlatforms #SlowMist #Cosine #ICRYPEXGlobal #Vulnerabilities #FundsSecurity #BlockBeats #CryptoSecurity #XPlatform

🚀 [ContractVulnerability] Shift in DeFi Security Approach Advocated by a16z Researcher

A senior security researcher at a16z Crypto, Daejun Park, has called for a shift in DeFi protocols from 'code as law' to 'specification as law,' advocating for a more principled security approach. According to ChainCatcher, Park suggests implementing standardized specifications and invariant checks to hard-code security measures, automatically reversing transactions that violate predefined rules. He notes that nearly all known vulnerabilities would trigger these checks, potentially preventing hacker attacks during execution.

A report from Slowmist highlights that hackers stole over $649 million last year through code vulnerabilities. Even established protocols like Balancer, which has been operational since 2021, suffered a $128 million loss due to code vulnerabilities in November last year. Developers are increasingly concerned about hackers using AI to find vulnerabilities.

The head of security at Immunefi points out that invariant checks could increase gas costs, potentially driving away users, and are not a cure-all solution. The co-founder of Asymmetric Research mentions that many vulnerabilities are challenging to write invariant rules for that can detect attacks without false positives.

#DeFi #Security #a16z #Blockchain #Vulnerabilities #InvariantChecks #Crypto #HackerAttacks #AI #SmartContracts #Slowmist #Immunefi #GasCosts #SpecificationAsLaw #ChainCatcher

🚀 Clawdbot AI Assistant Faces Security and Cost Challenges

Clawdbot, an open-source AI assistant, has rapidly gained popularity among developers for its advanced automation features across various messaging platforms. According to NS3.AI, despite its innovative capabilities, the platform has been found to have significant security vulnerabilities. These issues expose users to potential remote attacks due to default open network configurations. Additionally, high operational costs and extensive token usage have raised concerns, leading to calls for cautious adoption and improved security measures.

#Clawdbot #AIassistant #security #costchallenges #opensource #automation #messagingplatforms #vulnerabilities #remotetargets #tokenusage

🚀 DeFi Security Developments Highlight Mixed Outcomes Amid Market Uncertainty

Recent developments in decentralized finance (DeFi) security have shown a blend of successful hack recoveries and persistent vulnerabilities. According to NS3.AI, prominent firms are under fire for security shortcomings, while new funding commitments are being made to bolster Ethereum's infrastructure. Additionally, major cryptocurrency exchanges remain under scrutiny as market uncertainty continues in the wake of significant crashes.

#DeFi #security #hackrecovery #vulnerabilities #Ethereum #cryptocurrencyexchanges #marketuncertainty #funding #ETH

🚀 Dispute Arises Over Causes of October 10 Crypto Flash Crash

A disagreement has surfaced between cryptocurrency exchange executives and traders concerning the reasons behind the flash crash and mass liquidations on October 10. According to NS3.AI, the exchange's leader attributes the crash to Ethena’s USDe leveraged yield loop. In contrast, traders believe that thin liquidity or flawed market mechanisms were the primary causes. This debate underscores persistent worries about leverage and vulnerabilities in the market structure of crypto trading.

#crypto #flashcrash #liquidity #leverage #trading #cryptocurrency #marketstructure #ethena #usde #massliquidations #vulnerabilities #dispute

🚀 U.S. and Mexico to Develop Coordinated Trade Policy for Mineral Supply Chain Vulnerabilities

The United States Trade Representative (USTR) has announced that the U.S. and Mexico are planning to establish a coordinated trade policy aimed at addressing vulnerabilities in the supply chain of critical minerals. According to Jin10, this initiative is part of a broader effort to enhance the resilience and security of mineral supply chains, which are vital for various industries. The collaboration seeks to mitigate risks associated with supply disruptions and ensure a stable flow of essential minerals between the two countries. This move underscores the importance of international cooperation in maintaining robust supply chains amid global economic challenges.

#US Mexico #Trade Policy #Mineral Supply Chain #Vulnerabilities #Critical Minerals #Supply Chain Security #International Cooperation

🚀 OpenAI Launches EVMbench to Enhance Smart Contract Security

OpenAI has unveiled EVMbench, a benchmark aimed at evaluating AI models' capabilities in identifying, fixing, and exploiting vulnerabilities in smart contracts. According to NS3.AI, this initiative underscores the growing significance of comprehending smart contracts as AI agents may evolve into autonomous entities within crypto-native settings. The benchmark signifies progress towards incorporating AI-driven autonomous operations in blockchain ecosystems, with potential impacts on security and the infrastructure of decentralized economies.

#OpenAI #EVMbench #SmartContractSecurity #AIModels #Vulnerabilities #BlockchainEcosystems #DecentralizedEconomies #CryptoSecurity #AIAutonomy #Benchmark

🚀 AI Detects 92% of Real-World DeFi Vulnerabilities, Study Finds

A recent study reveals that specialized AI systems have successfully identified vulnerabilities in 90 DeFi smart contracts with a detection accuracy of 92%. According to Bitalk News, this is a significant improvement compared to GPT-5.1-based systems, which only detected 34% of vulnerabilities, involving a total of $7.5 million.

The specialized AI utilized domain-specific security methods, surpassing the limitations of traditional AI models. The study highlighted that the total losses associated with the smart contracts amounted to $228 million, with the vulnerabilities detected by the specialized system valued at $96.8 million.

These findings have raised concerns about the challenges AI poses in accelerating crypto-related crimes. Various studies indicate that AI can exploit smart contract vulnerabilities at a low cost, rapidly enhancing attack capabilities.

The research data and evaluation framework have been made available on GitHub as open-source, although the complete tools have not been released to prevent misuse.

#AI #DeFi #Vulnerabilities #SmartContracts #DetectionAccuracy #CryptoSecurity #BitalkNews #GPT5 #CyberSecurity #OpenSource #AIinCrypto

🚀 Digital Platforms Pose Risks to Young Men

Young men face significant vulnerabilities due to their exposure to digital platforms. Bloomberg posted on X, highlighting that these individuals are increasingly at risk through various online channels, including digital forums, social media, and gaming communities. The digital landscape presents unique challenges, making young men susceptible to various influences and potential harm. The need for awareness and protective measures is crucial to safeguard this demographic from the adverse effects of online exposure.

#DigitalPlatforms #YoungMen #OnlineRisks #SocialMedia #GamingCommunities #Vulnerabilities #Awareness #ProtectiveMeasures #OnlineExposure

🚀 China Warns of Security Risks in OpenClaw AI Agent

China's Ministry of Industry and Information Technology has issued a warning regarding security risks associated with the OpenClaw open-source AI agent. According to NS3.AI, the ministry's Cybersecurity Threat and Vulnerability Information Sharing Platform has reported the identification of new cybersecurity threats and vulnerabilities through recent monitoring efforts. The warning highlights the importance of addressing these security concerns to safeguard against potential risks.

#China #securityrisks #OpenClaw #AIagent #cybersecurity #NS3AI #vulnerabilities #cyberthreats #cybersecuritywarnings

🚀 Experts Warn of Persistent Security Risks in OpenClaw AI Agent

Experts from the China Academy of Information and Communications Technology have raised concerns about ongoing security risks associated with the open-source AI agent OpenClaw, also known as "Lobster." According to NS3.AI, despite a recent update aimed at addressing known vulnerabilities, the AI agent continues to face security challenges. The Ministry of Industry and Information Technology's cybersecurity threat and vulnerability information sharing platform had previously issued alerts regarding these security risks.

#OpenClaw #AIsecurity #ChinaAcademyofInformationandCommunicationsTechnology #Lobster #cybersecurity #vulnerabilities #NS3AI #securityrisks #MinistryofIndustryandInformationTechnology

🚀 China's MIIT Warns of Vulnerabilities in Apple Devices

China's Ministry of Industry and Information Technology (MIIT) has issued a warning regarding vulnerabilities in Apple devices operating on iOS versions 13.0 to 17.2.1. According to NS3.AI, attackers are exploiting these vulnerabilities through Safari, targeting users who open malicious links received via SMS, email, or compromised web pages. This exploitation can result in data theft and complete device compromise.

#China #MIIT #Apple #iOS #Cybersecurity #Vulnerabilities #DataTheft #TechNews

🚀 AI TRENDS | OpenClaw Vulnerabilities Addressed Following Discovery by 360 Intelligence

Three significant vulnerabilities in OpenClaw have been identified and reported by 360 Intelligence, targeting the core operational mechanisms of the AI system. According to Odaily, these vulnerabilities included one high-risk and two medium-risk issues, all of which have been officially fixed and publicly disclosed. The newly discovered vulnerabilities directly impacted the security of user devices, data, and accounts.

#AI #OpenClaw #Vulnerabilities #360Intelligence #Security #UserDevices #DataSecurity #AITrends

🚀 AI TRENDS | Anthropic Launches Project Glasswing to Enhance Software Security

Anthropic has announced the launch of 'Project Glasswing,' an initiative aimed at safeguarding critical software security worldwide. According to Foresight News, the project is powered by Anthropic's latest model, Claude Mythos Preview, which excels in identifying software vulnerabilities, surpassing most human experts and second only to top-tier technicians.

Anthropic has partnered with major companies including Amazon Web Services, Apple, Broadcom, Cisco, CrowdStrike, Google, JPMorgan Chase, the Linux Foundation, Microsoft, NVIDIA, and Palo Alto Networks. Together, they will utilize Mythos Preview to assist in detecting and rectifying system flaws. The model has already identified thousands of high-risk vulnerabilities, including some within major operating systems and web browsers.

#AI #Anthropic #ProjectGlasswing #SoftwareSecurity #ClaudeMythosPreview #Vulnerabilities #Cybersecurity #TechPartnership #AWS #Apple #Broadcom #Cisco #CrowdStrike #Google #JPMorganChase #LinuxFoundation #Microsoft #NVIDIA #PaloAltoNetworks

🚀 CertiK's AI Auditor Identifies 86.6% of Vulnerabilities in Web3 Security Test

CertiK has introduced AI Auditor, an AI-driven audit tool that demonstrated its capability by identifying 86.6% of vulnerabilities in a test involving 35 Web3 security incidents from this year. According to NS3.AI, the tool is designed to integrate security analysis directly into development workflows. CertiK aims to further expand the AI Auditor's application into developer tools, compliance systems, and institutional monitoring frameworks.

#CertiK #AIAuditor #Web3Security #Vulnerabilities #SecurityTest #NS3AI #AIinSecurity #Compliance #DeveloperTools #InstitutionalMonitoring

🚀 AI TRENDS | Anthropic Limits Claude Mythos Access Following Security Concerns

Anthropic has decided to restrict access to its Claude Mythos AI system to vetted cybersecurity groups. According to NS3.AI, this decision follows pre-release tests that revealed thousands of zero-day vulnerabilities and an 84% exploit success rate on Firefox 147's JavaScript engine. Anthropic's comprehensive 244-page system card highlighted that many existing cyber benchmarks were no longer adequately informative and disclosed significant evaluation oversights discovered late in the testing process.

#AI #Anthropic #ClaudeMythos #Cybersecurity #Vulnerabilities #ZeroDay #Firefox #JavaScript #Exploit #SystemCard #CyberBenchmarks #SecurityConcerns

🚀 AI TRENDS | U.S. Officials Meet Wall Street Leaders Over AI Cyber Risks

U.S. Federal Reserve Chair Jerome Powell and Treasury Secretary Scott Bessent held an urgent meeting with Wall Street leaders this week to address AI-driven cyber risks associated with Anthropic's Mythos model. According to NS3.AI, Anthropic reported that Mythos identified thousands of high-severity vulnerabilities and has restricted access through its Project Glasswing initiative. This meeting follows actions by U.S. agencies to limit the government's use of Anthropic products.

#AI #CyberSecurity #WallStreet #USGovernment #Anthropic #Mythos #ProjectGlasswing #FederalReserve #Treasury #Vulnerabilities

🚀 CMDSS Employee Accused of Misappropriating $46 Million in Seized Digital Assets

A federal indictment has been filed against John Daghita, an employee of CMDSS, accusing him of transferring approximately $46 million in seized digital assets from U.S. Marshals Service (USMS) wallets to wallets under his control. According to NS3.AI, Daghita was apprehended by French Gendarmerie officers on March 4 in Saint Martin. Prosecutors allege that these unauthorized transfers took place in December and January.

The case has reignited concerns previously raised by the Department of Justice (DOJ) inspector general and industry participants regarding the vulnerabilities in the USMS's crypto custody controls, procurement processes, and audit systems. These issues highlight the need for improved security measures and oversight in handling digital assets.

#CMDSS #JohnDaghita #USMS #digitalassets #crypto #seizedassets #fraud #misappropriation #federalindictment #FrenchGendarmerie #NS3AI #DOJ #cryptosecurity #vulnerabilities #audit #procurement #oversight