AI Dispatch — Always Be Rushing

JUL 17, 2026 · Moonshot AI, Inc. Magazine

Moonshot AI Releases Kimi K3, World's Largest Open-Source Model, Challenging US Rivals

Chinese startup Moonshot AI publicly rolled out Kimi K3, a 2.8 trillion-parameter open-source AI model, quickly achieving top rankings in front-end coding benchmarks. This release is seen as a significant development, narrowing the gap with advanced US systems like Anthropic's Claude Fable 5 and OpenAI's GPT-5.6 Sol in specific performance areas. The model's open-source nature allows developers to build upon its architecture.

READ DISPATCH ↗

Major AI Labs Ignite Price War with New, More Affordable Models for Enterprise

OPERATOR SIGNAL

JUL 14, 2026 · Kersai, Champaign Magazine

Major AI Labs Ignite Price War with New, More Affordable Models for Enterprise

SpaceXAI, OpenAI, and Meta released Grok 4.5, GPT-5.6 (Luna variant), and Muse Spark 1.1 respectively, igniting a fierce price war in early July. These new models offer significantly lower inference costs, with Luna priced at $1 per million input tokens and $6 per million output tokens. This aggressive pricing strategy aims for high-volume enterprise adoption and signals a market shift towards unit economics over raw benchmark scores.

READ DISPATCH ↗

OpenAI and Anthropic Advocate 'Subtracting' from Prompts for Enhanced Performance

PROMPT INTEL

JUL 13, 2026 · PromptArch, OpenAI, Anthropic

OpenAI and Anthropic Advocate 'Subtracting' from Prompts for Enhanced Performance

New official guidance from Anthropic (Claude Fable 5) and OpenAI (GPT-5.6) advises prompt engineers to simplify prompts by removing repeated instructions and examples. This 'subtraction' approach, also termed 'context engineering,' has demonstrated improved task performance and token efficiency, cutting costs by up to 67% in internal coding-agent evaluations. The focus shifts to providing essential context rather than overly prescriptive instructions.

READ DISPATCH ↗

Bank of America Reports Widespread AI Adoption Across 300+ Enterprise Use Cases

OPERATOR SIGNAL

JUL 20, 2026 · Bank of America, CIO Dive

Bank of America Reports Widespread AI Adoption Across 300+ Enterprise Use Cases

Bank of America's CEO, Brian Moynihan, announced that over 200,000 employees are utilizing AI-enabled capabilities, generating more than 400,000 prompts daily. The bank has over 300 approved AI use cases, including 114 generative AI applications, with 34 fully implemented in operations. This demonstrates significant real-world enterprise adoption for productivity tools, coding support, and advanced agentic workflows.

READ DISPATCH ↗

OpenAI Debuts ChatGPT Work, an Enterprise AI Agent Powered by GPT-5.6

MODEL UPDATE

JUL 20, 2026 · OpenAI, Forbes

OpenAI Debuts ChatGPT Work, an Enterprise AI Agent Powered by GPT-5.6

OpenAI has launched ChatGPT Work, a new enterprise AI agent specifically designed to automate complex workplace tasks. Powered by the GPT-5.6 model, this offering aims to challenge competitors in the enterprise AI space by focusing on execution and integrating AI into core business operations.

READ DISPATCH ↗

AGIBOT Unveils Four New Embodied AI Products at WAIC 2026 for Real-World Operations

MODEL UPDATE

JUL 20, 2026 · AGIBOT, PRNewswire

AGIBOT Unveils Four New Embodied AI Products at WAIC 2026 for Real-World Operations

AGIBOT, a leader in embodied AI and robotics, introduced four new products at the World Artificial Intelligence Conference 2026 in Shanghai: the A3 Ultra humanoid, X2 Edu, G2 Max industrial robot, and OmniHand 3 Ultra-M. These releases expand AGIBOT's portfolio across humanoid platforms, heavy-payload industrial task robots, and dexterous manipulation systems, emphasizing reliable integration into real operating environments.

READ DISPATCH ↗