AI

Datadog is expanding its AI-powered observability portfolio with new developer-focused tools aimed at helping engineering teams manage increasingly complex AI-driven applications and workflows.

Red Hat announced a major expansion of its developer portfolio this week with new tools focused on agentic AI development, signaling how enterprise software companies are rapidly adapting to the next phase of AI-powered software engineering. The announcement, made during the company’s Red Hat Summit event in Atlanta, introduced Red Hat Desktop and new capabilities within Red Hat Advanced Developer Suite.

AI innovation runs on data, but the industry has developed a bad habit of pretending that its data practices are simpler, cleaner, and more controlled than they really are. Behind the dashboards and governance slogans lies a growing discomfort: the more data organizations collect, the harder it becomes to justify how it’s used, secured, and shared. This tension sits at the heart of modern AI development - and it’s exactly where federated machine learning begins to change the story. 

Everyone has had that moment: a search or recommendation engine suggests something so oddly specific that it feels either magical or mildly unsettling. “How did it know I wanted that?” quickly becomes, “Should I be concerned that it knows that?” That tension - between delight and skepticism - is exactly where explainable AI enters the conversation.

OpenAI and ServiceNow have announced a strategic partnership aimed at embedding advanced AI agents directly into enterprise software development workflows. The collaboration is designed to help organizations automate complex tasks, accelerate development cycles, and improve operational efficiency across IT and business teams.

The central question is no longer just whether machines can write better tests than humans. It is whether AI systems - training pipelines included - can be built to respect privacy by design. Generative AI in testing relies heavily on data. Systems such as large language models, anomaly detection engines, and automated test case generators need vast and diverse datasets to function effectively. Yet this data dependency increasingly clashes with emerging privacy regulations that limit how persona

A few years ago, training an AI model felt like hosting an all-you-can-eat buffet. The more data piled onto the plate, the happier the algorithm seemed. Engineers collected logs, user clicks, purchase histories, and behavioral signals with enthusiasm. Storage was cheap. Compliance was an afterthought. Everyone went back for seconds. Then the regulators showed up. Suddenly, that generous buffet looked more like a liability. Data retention policies became urgent. Legal teams began asking

Software development is evolving faster than ever, and staying relevant requires continuous learning.

Artificial intelligence has an uncanny habit of sounding very sure of itself (kind of like this article ;-). It recommends a product, flags a transaction, denies a loan, or predicts customer churn - all with...

On December 1, 2025, industry leader NVIDIA announced a $2 billion investment in Synopsys, a major provider of electronic design automation (EDA) software, marking a significant strategic bet on

If algorithms were people, some of them would probably be asked to take a long, reflective walk. Imagine an AI hiring tool that keeps recommending applicants all named “Brad,” or a credit-scoring system that confidently insists anyone living on Maple Street has

Every QA engineer has experienced that sinking feeling: you fix one bug, run the automation suite, and suddenly a sea of red appears. Feel like a game of Wack-A-Mole, but this is no game!