Microsoft is reportedly tightening controls on internal AI token consumption, urging employees to choose models based on tasks and focus on business outcomes. The shift highlights growing enterprise pressure to balance AI performance with rapidly expanding infrastructure costs.

Consolidated AI chatbots raise order values up to 26% at Albertsons, while WalMart’s in-store AI use requires employees to further train AI models with business context.

After proving its custom silicon strategy with Maia 200, Microsoft is preparing a much larger Maia 300 push designed to support both AI training and inference.

ServiceNow adds six agentic security offerings designed to automate threat detection, governance, compliance, and remediation.

Google Cloud’s infrastructure strategy gives frontier AI lab Mirendil the flexibility and scale needed to speed up the AI research loop.

Skill Recorder creates a foundation for AI to understand workflows, recommend improvements, and ultimately automate workflows, for on-demand or scheduled use cases.

Palantir’s growth accelerated from 71% to 85% to 93% over three quarters as customer demand, expanding contracts, and a distinctive value-based pricing model fuel momentum.

Palantir’s rapid growth reflects an enterprise AI strategy centered on measurable customer outcomes, operational transformation, sovereignty, and deep long-term partnerships.

OpenAI has responded to the GPT-5.6-Sol security incident involving Hugging Face by strengthening cybersecurity safeguards, expanding its Daybreak program, and introducing GPT-5.6-Cyber for vetted defenders conducting advanced vulnerability research and security testing.

Salesforce scores a major AI win as the Pentagon approves its agents for autonomous tasks, reinforcing the security and governance credentials of its agentic infrastructure.

Enterprise AI’s next breakthrough will come from evaluating agents against real business workflows and continuously improving their performance with proprietary organizational knowledge.

Microsoft Research and Paige’s PRISM 2 combines nearly 2.5 million pathology images with clinical language, creating a versatile AI model for cancer detection, biomarker prediction, and research.

Palantir’s extraordinary 93% revenue growth and CEO Alex Karp’s unconventional philosophy could reshape enterprise software, emphasizing AI sovereignty, customer control of data, intensive partnerships, and compensation tied directly to measurable value creation.

Agentic reasoning loops promise better AI answers, but could dramatically increase token consumption and costs as enterprise adoption grows.

Palantir CEO Alex Karp is challenging traditional enterprise software with an approach that prioritizes customer data control, streamlined selling, and outcomes-based compensation.

Microsoft CEO Satya Nadella has confirmed plans for a Copilot super app that will unite chat, coding, Cowork and autonomous AI agents, creating a single flagship experience spanning both Microsoft’s consumer and commercial AI offerings.

AI vendors are moving toward consumption-based billing as rising token costs make traditional flat-fee and user-based licensing increasingly difficult to sustain.

Oracle is expanding its Google Cloud partnership by bringing Gemini models across its enterprise AI portfolio, strengthening its flexible, multi-model approach to enterprise AI.

ServiceNow CEO Bill McDermott is positioning the company as the enterprise AI Control Tower, combining governance, cybersecurity, workflow orchestration, integration, and automation to help customers manage the growing complexity and risks of enterprise AI.

AI vendors are increasingly turning to forward deployment engineers as enterprises struggle to move AI pilots into production and consumption-based models grow.

Bill McDermott is combining ServiceNow’s AI Control Tower, cybersecurity, and governance capabilities to give CEOs greater certainty and control as enterprise AI creates unprecedented risk and complexity.

Project Perception represents Microsoft’s push beyond reactive cybersecurity, coordinating red, blue, and green AI agents in a closed-loop system designed to continuously identify vulnerabilities, assess risks, implement fixes, and strengthen enterprise defenses.

Enterprise AI success increasingly depends on trusted, interoperable data that connects existing technology investments with autonomous processes and next-generation business models.

Oracle lands a major 10-year Pentagon contract worth up to $7 billion, strengthening its government credentials while creating a potentially significant stream of recurring revenue.

AI’s unprecedented growth is forcing even the world’s best-run companies to rethink traditional risk management and embrace massive, calculated bets on the future.

July updates extend the company’s commitment to make a range of AI models and choices available to users. Models from OpenAI and Anthropic have been added.

Microsoft, Oracle, Google Cloud, and AWS have amassed $2.3 trillion in contracted backlog, underscoring unprecedented demand across the booming cloud and AI economy.