.png){kind=logo}
Artificial intelligence agents could change how businesses purchase software, shifting enterprise technology from employee subscriptions toward automated, usage-based transactions. Paul Veradittakit, managing partner at Pantera Capital, argues that software itself may become the customer, creating opportunities for cryptocurrency payment infrastructure.
Software-as-a-service (SaaS) companies typically charge monthly fees based on employee seats. Their sales strategies depend on demonstrations, procurement teams, and interfaces designed for humans.
Autonomous agents operate differently. They can identify services, compare prices, and request computing resources through application programming interfaces without employees approving every transaction.
According to Veradittakit, this shift could reduce the importance of traditional dashboards while increasing demand for reliable, machine-readable services. Pricing could move toward API calls, computing consumption, or completed tasks rather than fixed subscriptions.
An AI research agent might purchase financial datasets, access analytical tools, and rent computing capacity within one workflow. Such transactions could involve numerous micropayments, making settlement speed, transaction costs and continuous availability important.
Cryptocurrency wallets and blockchain payment systems offer mechanisms for automated cross-border settlement and outside banking hours. However, widespread adoption requires safeguards against unauthorized spending, fraud and regulatory violations. Traditional payment networks could also support automated transactions.
The analysis identifies Fission as an example of an emerging ecosystem where agents can advertise services, request work, and exchange payments through decentralized infrastructure. Its model allows software to act as both buyer and supplier, potentially enabling continuous digital marketplaces. However, the source provides no verified transaction volumes, adoption figures or revenue data demonstrating commercial scale.
Developers may need to prioritize accessible APIs, transparent pricing, automated billing and verifiable machine identities. Agent reputation systems could help distinguish reliable providers from malicious services, while spending limits and approval controls remain essential.
Veradittakit's central thesis is that enterprise software purchasing could move from human-directed subscriptions toward autonomous procurement. The transition remains a projection, not an established industry outcome. Its success depends on trustworthy agents, practical payment systems and measurable business value.
Also Read: 10 Crypto Apps for Trading, Payments, and Everyday Banking in 2026