Apple nears deal to acquire computer vision startup Prompt AI

Apple is close to acquiring talent and technology from Prompt AI, a computer vision startup founded in 2023.

Prompt AI develops software that enhances home security cameras with object and activity detection.

The company raised US$5 million in seed funding from AIX and Abstract Ventures, and was founded by AI researchers Tete Xiao and Trevor Darrell, both linked to UC Berkeley.

Prompt executives told employees the deal will not fully repay investors, and staff not joining Apple will face reduced pay. Other companies like xAI and Neuralink also showed interest.

Prompt is shutting down its Seemour app, will inform users, and plans to delete user data for privacy.

Apple has historically focused on smaller acquisitions to build features for its devices, especially in smart home and computer vision technology.

🔗 Source: CNBC

🧠 Food for thought

Implications, context, and why it matters.

Apple’s pending acquihire hints at an AI push in home security, filling HomeKit gaps

Apple is reportedly building its first home security camera similar to Ring (Amazon’s home security brand), with facial recognition and infrared sensors that detect who is in the room 1. The Prompt AI acquihire (an acquisition to hire a startup’s team plus take its tech) would add computer vision that Seemour already uses to spot specific people or pets or objects.
HomeKit, Apple’s smart-home platform, includes HomeKit Secure Video that encrypts footage end to end and runs on-device analysis that sorts people or pets or cars 2. Prompt can produce text summaries and reply to questions about camera activity, which could feed a camera or a 7-inch hub expected in 2026 1 3 2. Apple also targets a smart doorbell with Face ID in late 2025 4.

Third-party HomeKit makers should watch for new computer vision Application Programming Interfaces (APIs) that enable richer automations

Camera vendors shipping HomeKit gear should watch the Worldwide Developers Conference (WWDC) for API updates that go beyond basic motion events, building on package detection 2.
iOS and smart-home app developers can use better vision features to build automations (pre-programmed actions triggered by events) that react to visual triggers, such as turning on lights when a delivery arrives or adjusting climate based on room occupancy 1 2.
Accessory makers should look to EnergyKit (a project focused on privacy-preserving home energy data processing) 5 plus Apple’s push for on-device processing with end-to-end encryption 6. Products that match these privacy rules will fit better inside Apple’s ecosystem.