Apple designs App Store framework for agentic AI amid revenue and security concerns
Internal efforts to create a controlled environment for agentic AI services reflect the company’s struggle to balance market inclusion with strict privacy and security standards ahead of its annual developer conference.
Apple is reportedly developing a new system to permit agentic AI services on the App Store, a strategic move designed to capitalise on growing developer and user interest while maintaining rigorous control over privacy and security standards. The potential policy shift addresses concerns that excluding such tools could marginalise the company in a rapidly expanding market, although permitting them introduces significant risks to revenue and security.
Agentic AI refers to software capable of taking active control over devices and their programmes. Apple has previously blocked "vibe coding" tools on the platform, citing violations of current policies. These tools have the potential to create original applications, which could divert users away from traditional app purchases and pose a threat to Apple’s bottom line. Additionally, there are concerns that such open-ended access could create loopholes for spreading malware or facilitating other malicious actions.
The decision to potentially allow agentic AI services involves balancing the need to avoid being excluded from a lucrative market against the risks of revenue loss and security vulnerabilities. According to reporting by The Information, Apple’s internal deliberations focus on designing a system that adheres to its strict standards while preventing the "freewheeling behaviour" observed in some agentic systems.
Examples of such instability include issues with OpenClaw, where agents have been reported to malfunction and delete user data, such as emails. Apple’s internal teams are reportedly working to mitigate these risks, ensuring that any permitted agentic services do not compromise the integrity of the user experience or the security of the ecosystem.
This development is expected to be contextualised within Apple’s broader AI integration plans for the upcoming Worldwide Developers Conference (WWDC). The company has faced challenges in integrating AI into its ecosystem while maintaining its reputation for privacy, and the outcome of these internal design efforts will likely be a key focus for investors and industry observers next month.
The specific technical details of the new system have not been publicly disclosed, and it remains unclear how Apple will enforce these standards without stifling innovation. The exact impact on Apple’s revenue model if agentic AI tools are permitted is also not quantified, leaving the market to speculate on the financial implications of this potential policy shift.
