Major enterprise software platforms, including Slack, Workday, and LinkedIn, are increasingly restricting how external AI agents can access their systems. This trend, driven by concerns over data privacy and competitive advantage, presents a structural conflict amid the rising demand for AI automation. Notably, Salesforce limited third-party access to its Slack data, and Meta updated WhatsApp’s API terms to exclude general-purpose AI chatbots. Operational risks have emerged, such as Google cutting access to its Antigravity platform due to high demand from AI agents like OpenClaw.
In contrast, the agent infrastructure market, exemplified by Arcade.dev’s ToolBench, is evolving rapidly to establish standardized connections between agents and platforms. With platforms and agents heavily dependent on each other for valuable data and workflow automation, the future likely involves differentiated access models for humans and AI agents. Ultimately, control over data access terms will shape the landscape of AI and enterprise software interactions.
Source link
