hodler

19 Feb 2026, 04:18

AI agents review smart contracts to identify and fix security issues that lead to crypto losses

Developers are now using AI agents to protect smart contracts that control billions in digital assets, following crypto hackers’ theft of over $3.4 billion from blockchain platforms in 2025. Rather than dozens of small thefts, 2025’s losses were concentrated in a few massive breaches, with just three major incidents accounting for nearly 70 % of the total value stolen. The most notable was the Bybit exchange hack , which alone siphoned off roughly $1.4 billion — one of the largest crypto thefts ever recorded OpenAI is working with Paradigm and OtterSec to test whether AI agents can detect vulnerabilities in real blockchain spaces using its EVMbench. AI agents review smart contracts to identify and fix security issues that lead to crypto losses Any error in smart contract code today will affect real money belonging to big and small investors, as these automated programs manage more than $100 billion in open-source digital assets. And after hackers stole over $3.4 billion from crypto platforms in 2025, developers can now see just how vulnerable the system is when attackers exploit weak code. Relying on human audits isn’t an option anymore because live contracts face new and evolving attacks that weren’t present during the audit process. Plus, it takes a lot of time and costs a fortune as security teams must review smart contract code before deployment. Instead of waiting for the next manual audit cycle that may come too late to stop an attack, developers are now turning to AI agents to continuously monitor live smart contracts. It takes AI agents less time to detect hidden code irregularities than people do, who may need days or even weeks, so frameworks like the EVMbench by OpenAI make more sense for developers. EVMbench uses AI agents in test environments to help developers understand how smart contracts may perform under real-world pressure before the actual deployment. The agents will first detect hidden vulnerabilities in code, fix the issue without breaking the contract’s function, and then try to exploit the weakness to drain funds if the problem persists. According to early results, AI agents are better at exploiting vulnerabilities than safely fixing them. People are now concerned that hackers could misuse AI-powered tools to exploit weaknesses in blockchain systems more efficiently than ever. AI agents can also create new security risks by helping hackers identify weaknesses in blockchain systems Machines are learning to break into weak contracts faster than ever before because current AI agent systems now succeed in exploiting more than 70% of vulnerabilities compared to earlier AI models with a less than 20% success rate. Attackers are now moving away from manual hacking methods and toward AI agents that scan large amounts of code and test different attack paths without direct human input. And as this trend continues, experts now say AI agents will soon be able to move funds, approve transactions, and manage financial tasks automatically on behalf of users. American technologist Jeremy Allaire said that billions of AI agents will soon use stablecoins to send and receive payments across blockchain networks. Founder and former CEO of Binance , Changpeng Zhao (CZ), also said crypto could become the native payment layer for AI-driven systems in the future. All these trends and predictions make AI agents more useful to both users and attackers, as they will soon interact with contracts directly in real financial environments where actual money is at stake. Industry leaders have even raised concerns about user safety. Managing partner at Dragonfly, Haseeb Qureshi, warned that many users still worry about sending funds to the wrong address or approving a harmful transaction by mistake through crypto transactions. To solve this problem, Qureshi proposed that AI-operated wallets could soon interact with the blockchain without users needing to understand the complex process involved. In this way, AI agents can assist in reducing human errors in audits and in protecting smart contracts by continuously monitoring systems. However, they can also increase the rate at which attackers discover vulnerabilities in the system, enabling exploits to scale much faster. This creates a security issue where AI systems developed to protect decentralized finance platforms could also be the most effective at attacking them if they fall into the wrong hands. The smartest crypto minds already read our newsletter. Want in? Join them .

18 Feb 2026, 23:30

Solana futures data shows panicked bulls: Will $80 SOL hold?

A drop in Solana’s dApp revenues, along with limited institutional and retail investor interest, adds vulnerability to SOL’s $78 support.

18 Feb 2026, 19:25

OpenAI Smart Contract Security: Revolutionary EVMBench System Transforms Blockchain Safety

BitcoinWorld OpenAI Smart Contract Security: Revolutionary EVMBench System Transforms Blockchain Safety In a groundbreaking development for blockchain security, OpenAI has officially launched EVMBench, a sophisticated benchmarking system designed to rigorously evaluate the smart contract analysis and security capabilities of artificial intelligence agents. This strategic initiative, developed in collaboration with leading crypto investment firm Paradigm, represents a significant advancement in automated blockchain security protocols. The announcement, first reported by Unfolded, signals a new era where AI systems will undergo standardized testing for their ability to detect vulnerabilities in Ethereum Virtual Machine-based smart contracts, potentially transforming how developers approach decentralized application security. OpenAI Smart Contract Security System Architecture EVMBench operates as a comprehensive evaluation framework specifically engineered to assess AI agents’ proficiency in analyzing smart contract code for security vulnerabilities. The system utilizes a diverse dataset of smart contract implementations, ranging from simple token contracts to complex decentralized finance protocols. Importantly, EVMBench measures multiple dimensions of AI performance, including vulnerability detection accuracy, false positive rates, and the ability to explain identified security issues in human-readable formats. The benchmarking platform incorporates real-world smart contract examples alongside deliberately vulnerable code samples, creating a robust testing environment that mirrors actual blockchain development scenarios. Transitioning from traditional security approaches, EVMBench introduces several innovative evaluation metrics. The system assesses AI agents’ capabilities across three primary categories: static analysis proficiency, dynamic behavior prediction, and exploit scenario identification. Furthermore, EVMBench evaluates how effectively AI systems can contextualize vulnerabilities within broader smart contract ecosystems, considering factors like contract interactions and protocol dependencies. This multi-layered approach ensures that AI agents demonstrate not just technical detection skills but also practical understanding of blockchain security implications. Paradigm Collaboration and Industry Impact The collaboration between OpenAI and Paradigm brings together complementary expertise in artificial intelligence and blockchain technology. Paradigm’s deep understanding of cryptocurrency ecosystems and smart contract vulnerabilities informed the development of EVMBench’s evaluation criteria. This partnership ensures the benchmarking system addresses real security concerns faced by blockchain developers and auditors. Industry experts anticipate that EVMBench will establish new standards for AI-powered security tools, potentially reducing smart contract exploits that have resulted in billions of dollars in losses across the cryptocurrency sector. Consequently, the introduction of EVMBench arrives at a critical juncture for blockchain security. The increasing complexity of smart contracts and the growing value locked in decentralized applications have created urgent needs for more sophisticated security solutions. Traditional manual auditing processes, while valuable, struggle to scale with the rapid expansion of blockchain ecosystems. EVMBench addresses this challenge by providing a standardized method to evaluate and improve AI-assisted security tools, potentially accelerating the development of more reliable automated auditing systems. Technical Implementation and Evaluation Methodology EVMBench employs a sophisticated technical architecture that simulates various blockchain environments and attack scenarios. The system evaluates AI agents through multiple testing phases, beginning with basic vulnerability detection and progressing to complex multi-contract interaction analysis. Each evaluation phase measures different aspects of AI performance, including: Code Pattern Recognition: Ability to identify common vulnerability patterns in Solidity and other smart contract languages Contextual Analysis: Understanding how vulnerabilities function within complete decentralized applications Exploit Prediction: Forecasting how attackers might leverage identified weaknesses Remediation Suggestions: Providing actionable security improvement recommendations Additionally, EVMBench incorporates temporal evaluation components, assessing how AI agents handle newly discovered vulnerability types and evolving attack vectors. This forward-looking approach ensures the benchmarking system remains relevant as blockchain technology and associated threats continue to develop. The platform’s design accommodates both general-purpose AI models and specialized security tools, creating a level playing field for different technological approaches to smart contract analysis. Blockchain Security Evolution Timeline The development of EVMBench represents the latest milestone in blockchain security’s ongoing evolution. The following table illustrates key developments leading to this innovation: Year Development Impact 2016 DAO Exploit Highlighted smart contract vulnerability risks 2018 Formal Verification Tools Introduced mathematical proof methods for contracts 2020 Automated Auditing Services Began scaling security analysis 2022 AI-Assisted Code Review Integrated machine learning into security workflows 2025 EVMBench Launch Established standardized AI evaluation framework This progression demonstrates how blockchain security has evolved from reactive measures to proactive, standardized evaluation systems. EVMBench builds upon previous innovations by creating measurable standards for AI performance in smart contract analysis. The system’s development acknowledges that as AI becomes more integrated into security workflows, standardized evaluation becomes increasingly essential for maintaining trust in automated systems. Industry Response and Future Applications Initial reactions from blockchain security professionals indicate cautious optimism about EVMBench’s potential impact. Security auditors note that standardized AI evaluation could help identify the most effective tools for different types of smart contract analysis. Meanwhile, blockchain developers anticipate that improved AI security tools will reduce development costs and time-to-market for secure decentralized applications. The benchmarking system may also influence insurance markets for decentralized finance protocols, as more reliable security assessments could lead to better risk pricing models. Looking forward, EVMBench’s architecture allows for expansion beyond its initial Ethereum Virtual Machine focus. The system’s modular design potentially supports adaptation to other blockchain environments and smart contract languages. This flexibility suggests that EVMBench could evolve into a universal standard for evaluating AI security tools across multiple blockchain platforms. Furthermore, the benchmarking data generated through EVMBench evaluations may inform academic research into AI capabilities and limitations in code analysis contexts. Conclusion OpenAI’s launch of the EVMBench smart contract security evaluation system represents a transformative development for blockchain technology safety standards. This collaborative effort with Paradigm establishes rigorous benchmarks for assessing AI agents’ capabilities in identifying and analyzing smart contract vulnerabilities. The system’s comprehensive evaluation methodology addresses critical needs in an industry where security failures have profound financial consequences. As blockchain ecosystems continue expanding, standardized AI evaluation through platforms like EVMBench will play increasingly vital roles in maintaining system integrity and user trust. The introduction of this benchmarking framework marks a significant step toward more reliable, scalable, and transparent security practices across decentralized application development. FAQs Q1: What exactly does EVMBench evaluate in AI agents? EVMBench evaluates AI agents’ abilities to detect, analyze, and explain smart contract vulnerabilities across multiple dimensions including detection accuracy, false positive rates, contextual understanding, and remediation suggestion quality. Q2: How does EVMBench differ from existing smart contract auditing tools? Unlike auditing tools that directly analyze contracts, EVMBench evaluates the AI systems that perform the analysis, establishing standardized performance benchmarks rather than conducting security assessments itself. Q3: Why is Paradigm’s involvement significant for this project? Paradigm brings extensive blockchain industry expertise and understanding of real-world security challenges, ensuring EVMBench addresses practical concerns faced by developers and auditors in cryptocurrency ecosystems. Q4: Can EVMBench be used with smart contracts on blockchains other than Ethereum? While initially focused on Ethereum Virtual Machine environments, EVMBench’s modular design allows for potential adaptation to other blockchain platforms and smart contract languages in future developments. Q5: How might EVMBench impact decentralized application development? By improving the reliability of AI-assisted security tools, EVMBench could reduce development costs, accelerate secure deployment timelines, and decrease vulnerability-related losses across blockchain ecosystems. This post OpenAI Smart Contract Security: Revolutionary EVMBench System Transforms Blockchain Safety first appeared on BitcoinWorld .

18 Feb 2026, 17:38

Shiba Inu’s AI Relationship Platform Sparks Web3 Expansion Talks

Shiba Inu’s lead ambassador has unveiled a new AI-powered relationship platform, while the project is also moving to protect users following a past security breach. The update links innovation with risk awareness across the SHIB ecosystem. Team members say the tool may expand into Web3 use cases over time. At the same time, developers warn users about rising phishing threats tied to a newly launched NFT initiative. AI Relationship Platform Eyes Web3 Integration During a Feb. 18 livestream, Shytoshi Kusama revealed plans for an AI-powered relationship platform focused on translation and compatibility. He described it as a personal initiative separate from his Shiba Inu leadership role. The platform aims to help couples detect patterns, friction points, and long-term compatibility risks before conflicts escalate. However, the Shiba Inu team sees broader Web3 potential. Lucie, a Shiba Inu team member, outlined five possible use cases connected to the SHIB ecosystem. She said the first could function as a DAO compatibility tool. The AI system could match co-founders, multisig partners, validators, or DAO teams to reduce internal conflict early. Lucie also explained that the platform may introduce a token-gated premium layer. This layer would unlock deeper AI insights using ecosystem tokens. She noted that this model could drive utility-based engagement across the community. In addition, she described a reputation and social signal system. The feature would link compatibility and communication insights to optional on-chain identity. According to Lucie, this could strengthen coordination within decentralized teams. She further suggested the tool could operate as an NFT or badge layer on Solana. The low-fee infrastructure could host collaboration or relationship milestone badges as social proof. Finally, she said the AI translator layer could serve as a mediator in governance discussions or advanced user support systems within ecosystem apps. Shiba Inu Issues Scam Warning Over SOU NFT Meanwhile, Shiba Inu has launched the SOU NFT as part of recovery efforts tied to the Shibarium hack last September. SOU, short for “Shib owes you,” exists as an on-chain NFT. The team describes it as a good-faith initiative to support impacted Shibarium users through payouts, donations, and occasional rewards. As the Shiba Inu SOU rollout continues, Lucie has issued a crucial scam warning. She stated on social media that scammers already operate fake SOU portals. She warned that phishing links now mirror official websites to drain user wallets. Lucie urged users to verify contract addresses and use hardware wallets. She advised the community to bookmark the official portal. She added that users should trust their instincts if something appears suspicious.

18 Feb 2026, 13:00

From 2016 hack to $150M Endowment: the DAO’s second act focuses on Ethereum security

Ten years after the famous hack, the DAO Security Fund has decided to stake the untouched ETH and use the yield to fund Ethereum security initiatives, honor claims indefinitely, and professionalize governance and key management.

18 Feb 2026, 12:45

Bitcoin Won Over Wall Street and Now It’s Paying the Price

Bitcoin’s Wall Street embrace was supposed to bring stability. Instead, it created a new vulnerability: dependence on American money that is now in retreat.