AgentLISA, the AI-powered security platform for blockchain infrastructure, announced today it has secured acceptance of three research papers at the 40th IEEE/ACM International Conference on Automated Software Engineering (ASE 2025), marking the first time a crypto-native organization has achieved triple publication at a single ASE conference.
The acceptance represents a significant achievement at a conference with a 20% acceptance rate, placing AgentLISA's research alongside work from leading academic institutions and established technology companies. The development underscores the company's strategy of translating peer-reviewed academic research directly into commercial security products.
Research Portfolio Spans Critical Security Vulnerabilities
The three accepted papers address fundamental challenges in smart contract security:
- •OpenZeppelin Vulnerability Analysis: Investigation into security flaws within the OpenZeppelin library, used by an estimated 60% of deployed smart contracts, and how these vulnerabilities propagate across the ecosystem
- •Access Control Benchmark: A comprehensive evaluation framework for tools designed to detect access control vulnerabilities, one of the most exploited attack vectors in decentralized applications
- •FaultSeeker Framework: An LLM-powered system for automated fault localization in blockchain transactions, reducing debugging time from hours to minutes
"Each publication directly informs our product architecture," said Dr. Izaiah Sun, Co-founder of AgentLISA. "The research isn't theoretical—it's immediately integrated into our multi-agent AI system, enabling detection of complex logic vulnerabilities that conventional tools miss."
Track Record at Top-Tier Venues
The ASE achievement extends AgentLISA's research presence across the industry's most selective top-tier conferences:
- •NDSS 2025: PropertyGPT paper earned Distinguished Paper Award at the Network and Distributed System Security Symposium, which attracted nearly 700 security researchers
- •ICSE 2024/2025: Two acceptances at the International Conference on Software Engineering, including work on exploitable bugs and autonomous cybersecurity frameworks
- •Usenix Security 2024: Research on detecting exploitable reentrancy vulnerabilities, a category responsible for over $300M in DeFi losses annually
The publication velocity is unusual for a commercial entity. Most crypto security firms publish sporadically; AgentLISA has maintained consistent presence at conferences typically dominated by university labs and Big Tech research divisions.
Commercial Translation: From Papers to Protection
The research-to-product pipeline has generated measurable traction. AgentLISA reports over 3,000 developer teams and 500+ premium subscribers across 14 blockchain networks. The platform integrates directly into VSCode, GitHub, and CI/CD pipelines, claiming 99% time reduction and 90% cost savings versus traditional manual audits.
Recent commercial initiatives include:
- •X402 Protocol Integration: Native support for machine-to-machine payments, enabling autonomous agents to programmatically request and pay for security scans. AgentLISA reached #4 globally on the x402scan leaderboard within weeks of deployment
- •Institutional Backing: $12M raised from Redpoint Ventures, UOB Venture Management, Signum Capital, NGC Ventures, and other institutional investors
Market Context and Competitive Positioning
The smart contract security market remains fragmented between traditional audit firms operating on weeks-long timelines and AI-native competitors focused primarily on runtime monitoring. AgentLISA's differentiation lies in comprehensive coverage: pre-deployment AI auditing combined with continuous monitoring, built on a foundation of peer-reviewed research.
With over $1.7B in protocol exploits during 2023, the company argues that deep research expertise enables detection of novel attack patterns before they're exploited in production.
About AgentLISA
AgentLISA is the first Agentic Security OS for Web3, making smart contract security accessible, affordable, and AI-native. Our multi-agent AI framework delivers enterprise-grade security analysis that is 99% faster and 90% cheaper than traditional audits, enabling developers to ship secure code with confidence.
Trusted by over 42,300 developer teams and 500+ premium subscribers, AgentLISA seamlessly integrates into development workflows through VSCode, GitHub, and CI/CD pipelines. Our platform supports 14+ blockchain networks and multiple programming languages, including Solidity and Rust.
Built on peer-reviewed research from top-tier conferences (NDSS, ICSE, Usenix Security, ASE), AgentLISA is setting the standard for autonomous security in the agent-to-agent economy.