AI Roundup — July 16, 2026

Here is a look at some of the notable AI industry developments from July 15–16, 2026.

Applied Computing Raises $20M to Build Foundation AI Model for Oil and Gas

Applied Computing has closed a $20 million Series A funding round with the goal of building a foundation AI model tailored specifically for the oil, gas, and petrochemical industry, according to TechCrunch. The company's aim is to provide operators with an AI model capable of representing an entire plant's operations — a significant vertical specialization effort in an industry that has historically been slower to adopt cutting-edge software tooling.

The raise signals continued investor interest in domain-specific foundation models as an alternative to general-purpose AI systems. Rather than adapting a broad model to industrial use cases after the fact, Applied Computing appears to be building from the ground up with the operational complexity of energy infrastructure in mind.

Microsoft Training Sales Teams to Position Its In-House AI as More Cost-Effective Than Competitors

TechCrunch reports that Microsoft is training its salespeople to present the company's in-house AI models as more efficient and cost-effective alternatives to models from OpenAI and Anthropic. The move reflects Microsoft's broader push to differentiate and monetize its own model development efforts, even as it maintains its well-known partnership with OpenAI.

This development is notable from a market structure standpoint: Microsoft is now actively competing in enterprise AI sales against the very companies whose models it has helped bring to market. For enterprise buyers evaluating AI vendors, the messaging suggests Microsoft intends to compete on both performance-per-dollar metrics and the integration advantages of its own stack.

Thinking Machines Releases First Open Model, Inkling

Thinking Machines has released its first publicly available model, called Inkling, according to TechCrunch. The release marks the company's first tangible public proof point after approximately a year and a half of building AI infrastructure largely outside of the public eye.

The company has positioned itself against what it describes as a "one-size-fits-all" approach to AI, suggesting that Inkling is designed with adaptability and specialization in mind rather than serving as a single general-purpose system. The open release of the model allows developers and researchers to examine and experiment with the approach firsthand.

Thinking Machines' emergence into the public-facing model space adds another entry to a growing list of organizations releasing open models, a trend that continues to shape how enterprises and developers evaluate their AI infrastructure options.

These developments reflect ongoing activity across AI verticals — from industrial specialization and open model releases to shifting competitive dynamics among major enterprise AI providers. Additional stories will be covered as further details emerge.