Smplify, a newly funded device-management startup, has raised $1 million in pre-seed capital and is positioning its product as a headless platform that runs entirely through APIs and a command-line interface.
AI agents raise governance concerns
Co-founder and chief operating officer Ben Greiner says the market for Apple device-management tools is expanding, yet most solutions still assume a human is continuously monitoring changes. “Device management still assumes a person is watching,” he told reporters.
That assumption is being challenged as AI agents begin to act on fleets at machine speed. Administrators have long automated routine tasks with scripts, but autonomous agents can now push updates, install software, or modify settings without human input.
According to a 2026 AI-ready governance report from OneTrust, 87% of organizations encourage the use of AI agents, while only 47% have clear controls in place. The gap creates a risk of unintended actions slipping through unnoticed.
A recent Gartner analysis predicts that 40% of enterprises will demote or retire autonomous AI agents by 2027 after governance gaps surface following incidents. The forecast suggests a growing market for tools that can audit and constrain AI behavior.
Industry leaders echo these worries. Jamf CEO Beth Tschida noted that AI adoption is inevitable and that building visibility and governance around it is more effective than trying to block it outright. Jamf introduced an AI Governance suite in July to address discovery, policy enforcement, and reporting.
In this environment, the question becomes who can approve the actions of an agent that can change thousands of devices in seconds. Greiner warned that “an agent with the right key can wipe thousands of laptops before anyone notices.”
Similar to the introduction of code-signing requirements for Windows updates a decade ago, the push for AI oversight reflects a pattern where rapid automation prompts new governance layers.
Smplify’s last-mile control plane
Greiner describes Smplify’s solution as a “last-mile control plane,” a final checkpoint that sits between an AI-driven decision and the actual execution on a device. The platform logs every request, the intent behind it, and the person or policy that authorizes the change.
The system integrates with Apple’s Declarative Device Management (DDM) framework, allowing administrators to define the desired state of devices. If a device drifts from that state—due to a missed update or a configuration change—Smplify can either automatically correct it or flag it for approval, based on predefined rules.
Many IT teams already use GitOps practices, storing configuration in GitHub and applying changes through pull requests.
Because Smplify is API-first, software vendors can embed native device-management capabilities into their own products without building a separate MDM solution from scratch. This approach could enable a broader ecosystem of AI-aware management tools across the enterprise.
Embedding governance into the workflow
Through its API-first design, the platform lets developers call management functions from existing CI/CD pipelines, keeping device control inside familiar tooling.
When a configuration change is proposed, the system records the originating request, the intended outcome, and the policy that must approve it before execution.
Versioned intent stored in a source repository creates a permanent history, allowing auditors to trace any alteration back to its source commit.
Risk mitigation and future boundaries
One concern highlighted by industry leaders is the difficulty of recovering from a mis-executed high-privilege script that runs with root access.
When code runs with full system rights, reverting changes may be impossible without a full device wipe, increasing downtime and cost.
To address this, the platform enforces rule sets that can block execution of commands requiring raised privileges until a human reviewer signs off.
Apple’s expanding controls over Apple Intelligence features give enterprises the ability to select which on-device AI models run on corporate hardware.
Preetham Gurram, a co-founder with a background at MobileIron, SOTI, and Apple, helped design the integration that leverages Apple’s zero-touch enrollment at scale.
