BitsLab Introduces AI-Agent Security Stack to Safeguard the Emerging On-Chain Agent Economy

SAN FRANCISCO, Dec. 11, 2025 /PRNewswire/ -- BitsLab, an AI security company dedicated to safeguarding digital assets, is introducing a unified security architecture for the emerging on-chain agent economy as autonomous AI agents begin taking on real financial operations. With machine-driven transactions becoming more common, BitsLab aims to address a core industry challenge: keeping autonomous economic agents safe.

From Audit Powerhouse to AI Security Infrastructure Provider

BitsLab's foundation is built on years of deep security work across ecosystems such as Sui, Aptos, TON, Solana, Linea, BNB Chain, Soneium, and Starknet. Through its audit brands—MoveBit, ScaleBit, and TonBit—the company has identified critical vulnerabilities across protocols, virtual machines, and cross-chain infrastructures. Combined with a strong record in global CTF competitions, this expertise supports BitsLab's evolution from a traditional audit firm into an AI-driven security infrastructure provider serving both organisations and everyday Web3 users.

Today, BitsLab secures more than 716 million users and over US$160 billion in on-chain value. Its products are powered by a multi-year data engine built from more than 200,000 vulnerability insights, extensive threat intelligence, and millions of malicious endpoint records—forming the backbone of both its consumer-facing BitsLab Safe and enterprise-focused BitsLab AI Scanner.

BitsLab Safe — AI Anti-Scam Protection for Everyday Web3 Users

BitsLab Safe provides real-time scam detection, safe browsing, and signature protection, helping C-end users navigate Web3 with confidence.

As user interactions increasingly involve AI agents and complex links, BitsLab Safe acts as a personal protection layer that catches threats before they cause financial loss.

BitsLab AI Scanner — Automated Vulnerability Detection for Enterprises

Designed for B-end clients, the AI Scanner accelerates smart-contract audits with dataset-trained models that automatically detect vulnerabilities, significantly improving audit speed and coverage for developers, enterprises, and protocols.

Why AI Agents Create New Security Risks

As AI agents gain autonomy—executing swaps, approving payments, managing portfolios—they introduce a new class of risks that traditional security models cannot handle. BitsLab defines this as Agentic Security: the protection of the decision-making processes of autonomous systems. Unlike smart contract security, which audits deterministic code, and AI safety, which focuses on alignment, agentic systems face threats such as prompt interference, behavioural drift, poisoned data sources, compromised training sets, and timing failures on chains with partial finality. These dynamics make securing autonomous AI fundamentally different—and far more complex—than securing conventional Web3 infrastructure.

Inside BitsLab's Agentic Security Stack

BitsLab's architecture revolves around two core modules:

1. AI-Agent Vulnerability Scanner

A pre-deployment tool that analyses prompts, reasoning logic and decision flows.

It evaluates whether an agent remains stable under adversarial input, using a dataset that blends symbolic vulnerabilities with behavioural patterns.

2. BitsLab Agentic Security Protocol (BASP)

A runtime 'trust gate' that evaluates whether an agent should be allowed to execute an on-chain action.

BASP uses:

• Identity & context validation
• Sandboxed simulations
• Dynamic whitelisting models

Designed to integrate with wallets, DEXs and machine-payment systems, BASP acts as the security checkpoint for autonomous economic agents.

Real-World Application: Securing Multi-Agent Payments with x402 and Questflow

A practical example of BitsLab's work can be seen in multi-agent payment systems built on:

• Coinbase's x402 protocol for machine-native payments
• Questflow's MAOP framework for coordinating distributed agents

BitsLab and Questflow have formed a long-term strategic collaboration to make agent-to-agent payments safer and more reliable.

Questflow's recent US$6.5M seed round shows rising confidence in multi-agent systems. Its upcoming Developer Platform will let developers package APIs directly as agents—expanding the environments where BitsLab's security stack is needed.

Here, BitsLab plays a critical role.

x402 and MAOP can achieve practical reliability only with a security layer that decides when an agent is allowed to spend funds.

BitsLab's Safe Agent models and security APIs provide:

• Risk scoring
• Anomaly detection
• Policy enforcement

These safeguards ensure that payments initiated by AI agents remain compliant, predictable and secure.

A Market Gap Emerging at the Intersection of AI and Web3

The industry still lacks a clear framework for answering the core question of agentic finance: when should a non-human system be permitted to execute an irreversible transaction? 

BitsLab believes the answer requires the co-evolution of three layers—payment primitives such as x402, multi-agent coordination frameworks such as MAOP, and security infrastructure that governs autonomous execution, including the BitsLab AI Scanner, Safe Agent, and BASP. As autonomous systems grow more capable and widespread, BitsLab warns that unregulated machine decision-making could become a significant systemic risk. A verifiable, data-driven security layer that protects both enterprises and everyday users will be essential for the safe scaling of the agentic economy.

BitsLab's Role Moving Forward

By combining enterprise-grade AI audit tools with consumer-focused anti-scam protection, BitsLab aims to become the default security standard for both organisations building AI agents and individuals interacting with them. 

BitsLab Safe keeps everyday users protected as AI-driven interactions become more common, while the BitsLab AI Scanner provides developers and enterprises with a stronger, more efficient security foundation. At the infrastructure layer, BASP and the broader Agentic Security Stack define the rules for safe and compliant autonomous execution. Together, these components position BitsLab at the centre of the emerging on-chain agent ecosystem—one where humans and autonomous agents will operate and transact side by side.

About BitsLab

BitsLab is an AI-driven security company focused on digital asset protection. It delivers an integrated solution of "audit services + AI security engine + security tools" for emerging Web3 ecosystems, helping developers and end users build, trade, and interact on-chain with greater safety.

Within its security framework, BitsLab combines BitsLab AI Scanner and BitsLab Safe to form a complete AI-powered protection system. BitsLab Safe, an AI-based Web3 security product, provides enterprise-grade defence by simulating transactions in real time, detecting scams and malicious contracts, and securing x402 payments and AI Agent operations through BitsLab's agentic security stack. BitsLab AI Scanner, built on a vulnerability and threat intelligence engine, performs intelligent auditing and risk detection to improve efficiency and reduce false positives.

BitsLab operates three sub-brands-MoveBit, ScaleBit, and TonBit-serving emerging ecosystems in Web3. The team offers professional audits and vulnerability research to keep core infrastructure secure during rapid iteration. Its researchers have won multiple international CTF awards and discovered critical vulnerabilities in many major ecosystems, driving meaningful improvements in industry-wide security.

Contact
Marketing Manager
Jason Lee
BitsLab
[email protected]

Photo: https://mma.prnewswire.com/media/2843403/BitsLab_AI_Agent_Security_Stack.jpg

SOURCE BitsLab