Andreessen calls AI the killer app driving crypto demand
Marc Andreessen, co-founder of Andreessen Horowitz, said artificial intelligence is becoming the defining use case for cryptocurrency, arguing that autonomous AI agents will drive demand for blockchain-based financial infrastructure.
Speaking on the Latent Space podcast, Andreessen described what he called a “great unification” between AI and crypto. He said AI agents will increasingly need financial capabilities, adding that early signs of this shift are already visible.
That transition is underway. Coinbase launched Agentic Wallets in February, a system designed specifically for AI agents rather than human users. The infrastructure enables autonomous spending, trading, and income generation on Base, its Ethereum layer-2 network. Its underlying payment protocol has already processed more than 50 million transactions.
MoonPay introduced MoonPay Agents, a non-custodial layer that allows AI systems to move assets across multiple blockchains, including Solana, Arbitrum, and Polygon. Meanwhile, Human.tech unveiled a wallet-as-protocol system that combines cryptographic oversight with spending limits for AI-driven transactions.
Other players are entering the space. Starchild, a research lab within the WOO ecosystem, has deployed autonomous AI agents on mainnet with wallets built on Ethereum-compatible infrastructure and Solana. These systems integrate language models, secure containers, and automated trading capabilities, allowing agents to execute transactions and transfer assets without human intervention.
The shift reflects a structural limitation in traditional finance. Banks and payment providers require identity verification that software agents cannot provide. Andreessen noted that some developers have already given AI agents access to bank accounts and credit cards as a workaround. He argued that crypto offers a more efficient solution.
Stablecoins are emerging as a preferred medium for agent-to-agent payments due to low fees and programmable compliance features. Analysts estimate that the global market for agent-driven commerce could reach between $3 trillion and $5 trillion by 2030, according to projections cited in the report.
The trend is drawing attention to Ethereum, which underpins much of the infrastructure used by these systems. Analysts say AI agents are already conducting transactions at scale on blockchain networks, raising questions about whether this activity will translate into sustained demand for block space.
Andreessen said the convergence between AI and crypto is no longer theoretical, but already unfolding across multiple platforms.
-
15:30
-
15:15
-
14:50
-
14:30
-
13:50
-
13:20
-
13:15
-
13:00
-
12:40
-
12:20
-
12:00
-
10:00
-
09:40
-
09:20
-
08:40
-
08:00
-
07:50
-
07:45
-
17:20
-
17:00
-
15:40
