Coinbase Launches Crypto Wallets For AI Agents
Agentic Wallets are designed to operate across networks, including the Ethereum layer-2 network Base, where agents can manage positions and execute strategies wherever opportunities exist. The approach envisions a future where agents autonomously optimize yields, rebalance liquidity, and deploy capital without requiring explicit, real-time approvals, provided permissions and controls are preconfigured by users. This marks a shift from AI assistants that merely advise to agents that act, according to Coinbase engineers Erik Reppel and Josh Nickerson in a Wednesday post announcing the development.
“The next generation of agents won't just advise - they'll act,” Reppel and Nickerson wrote, detailing plans for agents to perform a range of functions from monitoring yields across protocols to executing trades on Base and managing liquidity positions around the clock. They described a scenario in which an agent detects a more favorable opportunity at 3 a.m., rebalances automatically, and does so without explicit approval because user permissions and safety controls are already in place.
AI agents now operable on the Bitcoin Lightning NetworkBeyond Ethereum 's Base, Lightning Labs-the team behind Bitcoin 's Layer-2 Lightning Network-rolled out a new toolset enabling AI agents to transact on Lightning through the L402 protocol standard. The update also allows AI agents to run a Lightning node and manage a Lightning wallet containing native Bitcoin (BTC ) without accessing private keys. This development broadens the scope for autonomous financial activity on Bitcoin's network, providing a parallel pathway for agents to engage with programmable money at the base layer's second tier.
The push toward agent-enabled wallets comes alongside broader industry activity. Crypto CEO Kris Marszalek announced ai, a platform intended to let users create personal AI agents to perform everyday tasks on their behalf. The capability ranges from managing emails and scheduling meetings to canceling subscriptions, shopping tasks, and even trip planning. Marszalek described a spectrum of tasks that AI agents could handle, illustrating how these tools might eventually operate as your digital proxy across daily routines.
Why crypto leaders are embracing agentic AIIndustry executives have long warned that AI could redefine how value is exchanged online. In late January, Circle CEO Jeremy Allaire suggested billions of AI agents could transact with crypto and stablecoins for everyday payments within three to five years. Former Binance CEO Changpeng Zhao has echoed a similar sentiment, arguing that a native currency for AI agents is likely to be crypto, capable of supporting tasks from purchasing event tickets to paying restaurant bills. These public statements reflect a shared belief that programmable money and autonomous agents will converge to enable more fluid, real-time financial interactions.
At a higher level, the convergence of AI with decentralized finance and payments ecosystems is driving experimentation around agent autonomy. Google 's recent Universal Commerce Protocol, announced in January, is designed to power agentic commerce by enabling agents to initiate transfers on a user's behalf, with Google Pay acting as the default payment handler for USD-denominated transactions. The protocol signals a broader push in the tech sector to enable AI-driven commerce that can operate across apps, devices, and payment rails without constant human oversight.
As these capabilities mature, momentum in the space is likely to hinge on two dimensions: the robustness of autonomous decision-making and the security of permissioning and governance models. Agentic Wallets must balance the convenience of automated actions with safeguards to prevent unintended risk exposure. The ongoing conversations around risk controls and regulatory alignment will shape how broadly such wallets are adopted by retail and institutional users alike.
Market contextThe emergence of autonomous wallets sits within a broader cycle of increased on-chain programmability and the maturation of smart contract-enabled finance. As liquidity provision, yield optimization, and creator economy participation become more automation-friendly, the appetite for self-operating agents grows among developers and institutions alike. The convergence of AI tooling with established networks like Base and the Lightning Network underscores a dual-track approach: one path leverages scalable, smart-contract-enabled ecosystems, while the other emphasizes fast, low-friction payments on Bitcoin's secondary layer. Regulatory clarity and ETF-related flows in traditional markets are likely to influence how aggressively capital participates in these early-stage, automation-centric use cases.
Why it mattersAgentic Wallets represent a tangible step toward programmable money that can autonomously allocate capital, monitor risk, and adjust exposure across multiple protocols. If successful, the approach could reduce the overhead of manual trading and portfolio management, enabling more people to experiment with sophisticated strategies without in-depth technical know-how. The ability to manage DeFi positions and pay for compute or data access autonomously also has implications for developers building AI-powered financial tools, potentially accelerating product development cycles and new business models in the crypto space.
The integration with Bitcoin's Lightning Network adds a separate layer of significance. By enabling AI agents to transact via L402 on Lightning and hold a Lightning-compatible wallet, the ecosystem expands the set of on-chain and off-chain rails that can be orchestrated by autonomous programs. This broadens practical use cases for AI agents-from micro-payments to cross-network arbitrage-while testing the limits of permissioned automation and the user controls that balance safety with convenience. Taken together, these developments suggest a future in which agents operate across multiple rails with varying latency, fees, and settlement characteristics.
For users and builders, the key takeaway is a shift in how wallets are used and who controls them. Agentic Wallets place agency in the hands of AI-enabled programs, but with computerized governance that requires explicit permissions ahead of time. The risk-management framework around such permissions will be critical to its sustainable adoption, particularly as public enthusiasm for automation intersects with concerns about security and misuse. The coming months are likely to reveal the first generation of real-world deployments and decision-making heuristics that will define the role of agents in everyday crypto activity.
What to watch next- Expansion of Agentic Wallets beyond Base to other Ethereum layer-2s and compatible networks, including any developer updates from Coinbase. Tracking adoption and volume on the x402 payments protocol, including any reported milestones beyond the 50 million transactions already noted. Broader deployment of AI agents on Bitcoin via the Lightning Network using L402, and the integration of wallets with Lightning node operations. Progress and practical traction for ai by Crypto, including user adoption metrics and featured autonomous tasks. Further details on Google's Universal Commerce Protocol and collaboration milestones that enable agent-initiated transfers and payments in real-world settings.
- Coinbase: Introducing AgentKit - developer-facing overview and the roadmap for embedding wallets into autonomous agents. Coinbase Developer Platform status updates on AgentKit and Agentic Wallets deployment. Lightning Labs: L402 protocol standard enabling AI agents to transact on Lightning and manage Lightning-enabled Crypto: ai platform launch and its scope for personal AI agents performing daily tasks. Google: Universal Commerce Protocol and Agent Payment Protocol 2 for agent-enabled transfers in commerce.
Coinbase's public framing of Agentic Wallets as a step toward“agents that act” follows a broader wave of AI-powered automation across crypto layers. The combination of AgentKit, x402, and multi-network reach-spanning Base and the Lightning Network-provides a multi-faceted testbed for autonomous financial activity. Investors and builders will be watching for evidence of sustainable user authorization models, transparent risk controls, and clear metrics around automated yield optimization and liquidity management. As the ecosystem experiments with agent-based transactions, market participants will assess whether these autonomous wallets can reliably operate without compromising security or user intent.
Risk & affiliate notice: Crypto assets are volatile and capital is at risk. This article may contain affiliate links.
Legal Disclaimer:
MENAFN provides the
information “as is” without warranty of any kind. We do not accept
any responsibility or liability for the accuracy, content, images,
videos, licenses, completeness, legality, or reliability of the information
contained in this article. If you have any complaints or copyright
issues related to this article, kindly contact the provider above.

Comments
No comment