Author: Christian, a16z Frontier GTM
Translator: Gu Yu, ChainCatcher
Our work in the go-to-market strategy team at Frontier allows us frequent interactions with many leading companies at the AI infrastructure layer, covering everything from computing power suppliers and managed inference businesses to development teams building agents. As collaboration deepens, we have been exploring how cryptographic technology can assist these companies in their development.
Given that stablecoins have the potential to make payment processes faster, cheaper, and easier to manage, I initially thought payments would be the clearest entry point. However, computing power and inference suppliers are focusing their attention on growth and the fundamentals of business operations: ensuring power reserves, deploying online, and managing profit margins. For teams building and deploying agents, the frequency of mentions of reliability and controllability far exceeds that of payments: What can the agent access? What actions has it been authorized to perform? How should the company verify what actually happened?
Payments still represent a promising opportunity for crypto x AI. But these conversations reveal another possibility: at least in the short term, verifiability is a more critical application scenario for blockchain. As companies entrust agents with greater responsibilities, they need a mechanism to determine who authorized an action, what restrictions apply, and whether that authorization is still valid.
For entrepreneurs, a more direct opportunity lies in helping companies maintain tight control over agent behavior while delegating more work to them. The rich experience of cryptographic teams in signing and authorization delegation opens the door for them to enter this market.
Delegating More Responsibilities to Agents
Delegating more responsibilities to agents requires reliable performance and clearly defined limitations on agent privileges. If an agent is constantly messing up tasks, simply relying on permission controls won't solve the problem. However, when an agent performs exceptionally well and provides practical value, companies still need to decide what it can access, which actions require approval, and how to revoke its access rights. This is where the expertise of cryptographic teams in signing and authorization delegation truly shines.
Before a company expands the responsibilities of an agent, it must clearly define what the agent can access, what actions it can perform, and how to enforce those limitations—especially in this new world where interfaces (and user sense of control) are changing. This principle applies whether the agent is reading corporate information for ordinary consumers or modifying production environment systems for developers.
Imagine a company that uses an agent provided by an external software vendor to investigate a production incident. The agent can read logs and propose fixes, but must obtain approval before writing to and modifying the system. The company grants the permissions, the vendor runs the agent, and the system receiving the request must verify whether the action is permitted. However, a survey conducted in January 2026 among 150 IT and security decision-makers showed that 69% of respondents stated that security concerns were slowing the adoption of agents. Only 27% of respondents believed that their existing identity systems could fully govern agents in large-scale scenarios. Major concerns included data breaches and excessive permissions.
Meanwhile, research on deployed agents also indicates that there is substantial work to be done before companies grant them greater autonomy. In the paper "Measuring Agents in Production" (which explores why institutions build agents, how they build them, how they evaluate them, and the main development challenges they face), reliability was listed as the top developmental challenge. Teams often restrict agents to short workflows while keeping human intervention in the process. Jellyfish’s latest engineering data also shows that while agent participation is increasing, the ability to move tasks from requirements specification to code merge requests (PR) without human intervention remains relatively rare.
Code-writing agents are another example: they may generate very useful code modifications but still require human approval for each deployment. Companies need to define which systems it can change, which actions require review, and how to revoke its access rights. Better control mechanisms will enable companies to confidently delegate more work to these two types of agents.
Empowering Companies to Delegate
When agents work across multiple companies, the issues become more complicated. In the earlier mentioned production incident example, the company grants permissions, the vendor runs the agent, and the receiving system must enforce restrictions. If there is a verifiable record of agent permissions, both the company and the vendor have a common foundation to verify whether an action was authorized.
Existing identity systems and application permissions can already address some aspects of this problem. For cryptographic technology or related companies, the entry point lies in leveraging their team's experience in signing, authorization delegation, and user-controlled account management to make authorizing, executing, revoking, and auditing delegations among these systems easier.
Cryptographic teams have begun applying this experience to the challenges faced by enterprises: enabling software to take actions without granting it unlimited permissions. In the Turnkey Wayfinder case study, Parallel Studios implemented a session-based signing mechanism that allows users to authorize multiple transactions at once without needing to sign each individually. According to the case study, this method preserves users' custody over their wallets.
Wayfinder applies this capability to on-chain transactions. What interests me and the company teams I interact with is enabling software to act without granting it unrestricted powers. The greater opportunity lies in allowing agents to perform more work within companies' enforceable and revocable permission limits while retaining records of all authorized content.
When independent parties need a shared permission record that is not controlled by any single entity, or when the assets being controlled are already on-chain, blockchain can create value. However, parties must still reach a consensus on who has the authority to grant permissions, what operations the permissions allow, and how the revocation mechanism works—which is precisely the original design purpose of blockchain.
For off-chain applications, companies can record the permissions of agents and their expiration times, along with any revocation records, on-chain. This registration system can run alongside existing off-chain identity and application systems, which will cross-reference the on-chain system before allowing an action to verify the agent's signing request. Both the company and the vendor can verify what has been authorized, while the receiving system continues to be responsible for enforcing these restrictions.
As many of these agents operate off-chain, reaching enterprise customers means that any built-in control mechanisms must seamlessly integrate into the existing systems these customers are already using. For cryptographic teams, this often means extending products beyond on-chain applications. To bridge this go-to-market gap, these teams need to demonstrate how their experience can assist companies in empowering agents with greater responsibilities while not granting them more access than necessary. The most pragmatic testing criterion is: can target customers let agents handle a task that was previously impossible to delegate due to permission issues?
Simply telling companies to "trust their agents" is hardly a compelling reason to purchase your product. Instead, entrepreneurs should first target specific tasks that enterprises want to delegate but cannot do so due to permission or approval requirements. Then demonstrate how your product removes this barrier. Based on this, identify why this improvement is worth adopting a new system—does it provide sufficient value? Can it improve the company's bottom line, or help it explore new business areas?
Who Will Pay for This?
In exploring enterprise-level go-to-market strategies, another crucial step is identifying the right buyers:
For companies deploying agents internally, the buyer is likely to be the identity or security engineering lead, with the CISO being a potential sponsor. The problems they hope to solve may include rampant agent permissions, slow revocation processes, or investigations requiring behavior tracking across multiple systems. You should refine your pitch around a specific deployment scenario that your product can unlock, demonstrating how it will integrate seamlessly with the company's existing security system.
For software companies selling agents, the entry point should be the CTO or the engineering lead for the agent platform. Ask them how much authorization development work is needed for each deployment targeting specific customers, and whether this work is slowing down product release timelines or consuming engineering resources. Show them how your product can meet these customer needs with less customization work and speed up deployment progress.
In either case, founders must understand who is maintaining the current alternatives, what the costs are, and whether they are more likely to resolve the issues themselves or willing to pay existing vendors to handle them.
Existing identity vendors have an inherent advantage because their systems are already in place. Even if a new product offers better control features, target customers will always weigh this benefit against the cost of integrating and maintaining another tool. (Some salespeople liken this internal resistance to countering gravity.) Founders must understand the shortcomings of existing systems and why customers would choose your product over merely extending their current system. The hurdles to overcome are not just filling necessary gaps but also creating an enhancement that makes its value clearly outweigh the efforts involved in introducing a new tool. Unlocking a significant agent implementation might justify the substantial integration work; however, a small operational improvement must feature characteristics that make it easy to implement. Start with a very clear workflow, and take on as much of the integration work as possible yourself, so the target enterprise doesn’t have to go through great lengths to get everything sorted.
If these conversations reveal an opportunity gap, start small and materialize the value. Enable the agent to perform a specific operation under executable and revocable permissions from the customer, and return the records to the customer's existing auditing tools. You can validate through pilot projects whether this approach saves customers integration work or enables them to delegate tasks that would otherwise require human review while preventing unauthorized actions.
To assess potential customer opportunities, you or your sales team need to ask the following questions:
What tasks do customers want to delegate, and what is currently preventing them from doing so?
Who is maintaining the current alternatives, and what are the costs involved?
What improvements would justify the introduction of a new system?
What will it cost the customer to adopt this solution, and how much of that can you take on?
What functionalities does your solution provide that existing customer tools cannot or cannot easily achieve?
You should clarify the answers to these questions before engaging with customers, not afterward.
Next, Let’s Talk About Payments
New payment infrastructure will only be truly adopted when the barriers to achieving the customer’s desired goals have been eliminated. Simply increasing agent adoption rates does not tell us which workflows need new payment infrastructure. The precise needs depend on specific workflows:
An agent purchasing unknown API access may require an immediate payment method.
A code-writing agent operating within an enterprise’s existing subscription framework may not encounter any payment barriers at all.
Computing power suppliers may benefit from a different settlement method, but they also have more urgent operational challenges to prioritize.
Optimizing payment will only become a primary task when existing systems obstruct customers from achieving their desired goals. For example, the agent may need to manually set up an account before purchasing access to a service; a certain billing model may render small purchases impractical; or the settlement process may bring significant operational burdens. These are all valid reasons to consider adopting different payment systems.
Payments also require control mechanisms. For example, AP2 utilizes cryptographically verifiable authorization instructions (mandates) to record user authorizations for operations performed by agents, including those carried out in the absence of the user. However, customers may indeed need a more comprehensive control mechanism and not necessarily a brand-new payment method.
Simply increasing transaction volume is not sufficient to establish demand for new payment infrastructure unless you clarify what customers are actually purchasing and why existing payment methods cannot meet those needs.
***
As we enter the era of the agent economy, the core question is not how to integrate stablecoins into AI businesses. Rather, what will allow these companies to confidently delegate more responsibilities to agents? Founders in the crypto industry and other frontier fields should engage more with these teams to attempt to understand and bridge this gap. Selling groundbreaking products to enterprises follows the same principle: you must always start from where the customer currently is, not just where you want them to reach.
免责声明:本文章仅代表作者个人观点,不代表本平台的立场和观点。本文章仅供信息分享,不构成对任何人的任何投资建议。用户与作者之间的任何争议,与本平台无关。如网页中刊载的文章或图片涉及侵权,请提供相关的权利证明和身份证明发送邮件到[email protected],本平台相关工作人员将会进行核查。