GPT-5.6 Takes Over Work. Agents Raise $100M and Hunt Ethereum Bugs

In partnership with

Master Claude AI (Free Guide)

The professionals pulling ahead aren't working more. They're using Claude.

Our free guide will show you how to:

• Configure Claude to be the perfect assistant

• Master AI-powered content creation

• Transform complex data into actionable strategies

• Harness Claude’s full potential

Transform your workflow with AI and stay ahead of the curve with this comprehensive guide to using Claude at work.

Get The Free Guide

OpenAI Turns ChatGPT Into a Full Work Agent With GPT-5.6

OpenAI launched ChatGPT Work, an agent that can act across connected apps and files, remain on complex projects for hours, and produce finished spreadsheets, presentations, documents, and web apps. The company says Codex now has more than 5 million weekly users, including over 1 million people using it outside software development. Scheduled Tasks can monitor websites, Slack, email, and other connected systems, while the new desktop app adds a built-in browser and computer use across local applications.

The accompanying GPT-5.6 family—Sol, Terra, and Luna—introduces an “ultra” mode that coordinates multiple agents, an API multi-agent beta, a 92.2% BrowseComp score, and pricing ranging from $1 to $5 per million input tokens. GPT-5.6 is also becoming the preferred model inside Microsoft 365 Copilot, positioning OpenAI to compete directly for end-to-end enterprise work rather than isolated chatbot interactions.

Meta Enters the Agentic Model Race With Muse Spark 1.1

Meta released Muse Spark 1.1, a multimodal reasoning model designed specifically for coding, tool use, computer operation, and long-running agentic workflows. The model supports a 1-million-token context window and can manage its own context by retrieving older work and compacting completed steps. It can act as a primary orchestrator that plans projects and delegates tasks across parallel subagents, or operate as a specialized subagent within another system.

Meta also opened a public preview of its new Meta Model API, with reported pricing of $1.25 per million input tokens and $4.25 per million output tokens. Strategically, the release moves Meta beyond open model distribution and into the paid enterprise model-and-agent infrastructure market dominated by OpenAI, Anthropic, and Google.

An AI Agent Runs Point on Lyzr’s $100 Million Fundraise

Enterprise agent startup Lyzr used an in-house AI agent to handle much of the investor process for a proposed $100 million Series B at an approximately $500 million valuation. The agent reportedly answered questions from more than 130 investors, drafted dozens of investment memos, and monitored which presentation slides attracted the most attention.

Lyzr says the campaign generated roughly $400 million in potential interest from Silicon Valley, Middle Eastern, and financial-sector investors. The round was still being assembled on July 9, no lead investor had been disclosed, and the company stressed that the agent opened conversations rather than closing final commitments.

The exercise is strategically significant because Lyzr turned a high-stakes internal workflow into a live demonstration of the enterprise agents it sells.

Ethereum’s Agent Fleet Finds a Real Protocol Vulnerability

The Ethereum Foundation disclosed that coordinated AI agents had found real vulnerabilities in protocol-related software, including a remotely triggerable panic in libp2p’s gossipsub fixed as CVE-2026-34219. The security team runs multiple agents in parallel, with shared repository state replacing a central orchestration process.

Agents take on specialized roles covering reconnaissance, vulnerability hunting, coverage gaps, validation, and deduplication. Every candidate must include a reproducible artifact and survive independent checks because most initial reports are incorrect, duplicated, or unreachable in real configurations.

Ethereum’s main conclusion is strategically important for agentic security: finding possible bugs is becoming cheap, while rigorous triage and verifiable evidence are now the real bottlenecks.

Think You Know What AI Does Next?

Which model leads the next benchmark? Which AI lab ships the next major breakthrough?

Kalshi lets you trade on real-world AI and technology events as the industry moves. If you follow launches, model updates, and benchmarks closely, put that knowledge to work.

Trade $25, Get Up to $500

_{Bonus credit varies from $15 to $500. Terms apply.}

🗺️ 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 →