Waymo Put Agent in a Car, Klarna Put One in Commerce, SOMPO Put 30,000 People on It

In partnership with

The Latest Agentic Development

Waymo Tests Gemini as In-Car Agent for Robotaxi Fleet

Waymo is testing Google's Gemini AI chatbot as an in-vehicle assistant in its autonomous taxi fleet, with a 1,200+ line system prompt discovered in app code. The agent handles passenger queries, controls cabin features (lighting, temperature, music), provides personalized greetings based on trip history, and maintains clear separation from the autonomous driving system. This represents a live agent deployment across thousands of production vehicles.

Klarna Launches Agentic Product Protocol

Klarna released the Agentic Product Protocol, an open standard enabling AI agents to instantly access structured feeds of over 100 million products across 12 markets. The protocol standardizes product data for real-time comparison, recommendations, and pricing eliminating the need for merchants to reformat data or use intermediaries. Merchants connect via API to make inventory discoverable to any compliant agent.

SOMPO to roll out “SOMPO AI Agent” to ~30,000 employees

Sompo Holdings announced it will deploy AI agent tools to ~30,000 employees across domestic group companies starting January 2026, calling it (as of Dec 2025) the largest-scale domestic deployment within a single corporate group. The program focuses on agent-assisted internal search/summarization, “deep research,” meeting minutes, and analysis—plus custom agents integrating internal data/systems—explicitly positioning agents as a lever for broad process redesign and productivity gains.

Shippo launches agentic shipping: “Shippo MCP” turns plain-English chats into booked shipments

Shipping platform Shippo launched Shippo MCP, positioning it as an agentic layer on top of its multi-carrier API so assistants/devs can execute real shipping tasks via natural language (book/manage/adjust shipments). The key enterprise implication is governance: as “agentic shipping” touches cost, compliance, and liability, adoption will hinge on observability, human-in-the-loop controls, and auditable decision logs.

Tuya launches “Hey Tuya” super assistant built on a multi-agent architecture (physical AI angle)

Tuya disclosed “Hey Tuya,” a consumer-focused “Super AI Life Assistant” described as multi-agent and designed to move beyond reactive commands into proactive, context-aware routines—an important tell that consumer IoT players are trying to make “agentic” ambient and device-native. Tuya also framed this as part of a broader developer ecosystem (its AI agent dev platform + open-source framework), emphasizing scale via its developer base.

HodlHer Raises $1.5M for AI Agent Web3 OS

The capital will fuel development of its “HodlOS” platform – focusing on emotional AI interaction, multi-agent systems, and an on-chain agent economy – to enable autonomous agents that understand users’ emotions and act safely on their behalf in Web3.

What to Watch this Week

 🎙️AI on Fire: Real Builders. Real Heat.

Stories from founders building AI Agents. Explore all episodes here

⚔️ AArena: The Battleground for AI

Stop demo-hopping. One workspace. Every agent. Real results.

👉 Enter the Arena | 👉 See the Battleboard

🏆 The Leaderboard Never Sleeps

The global ranking of AI agents is shifting every day. Who’s on top? Who just dropped?

👉 Explore the Leaderboard →

🗺️ The Map of AI Agents (Live & Growing)

We’re charting the entire AI agent ecosystem — thousands of options across categories.
Your next agent is already on the map.

👉 Explore the Landscape Map →