Microsoft uses AI to find Windows flaws sooner and publish fixes faster
Why it mattersIf you run Windows across your company, faster published fixes mean your IT team has less time to install updates.
Why it mattersIf you run Windows across your company, faster published fixes mean your IT team has less time to install updates.
What happened. Microsoft says it is using AI, including a code scanner called MDASH that runs several AI models and specialised agents, to find and validate Windows vulnerabilities earlier, with human teams reviewing the results. It also says it will publish details of each fix promptly and is improving how organisations roll out updates.
Why it mattersIf your marketing team pays per second for AI video, each clip now costs less and runs five seconds longer.
What happened. Google's Gemini Omni 1.2 Flash now generates and edits video up to 15 seconds, up from 10, and costs 30% less than version 1.1, at $0.07 per second for 720p. It is generally available through the Gemini API, Google Cloud and Gemini for business.
Why it mattersIf you run or buy AI systems, memory prices push your costs up, and software efficiency now decides how much you pay.
What happened. NVIDIA says GPU memory is under growing strain as bigger AI systems, extended inputs and task-running assistants demand more of it, while supply tightens and prices rise. Its answer is to design chips, networking, storage and software together so each byte of memory does more work.
Why it mattersMerge applies OpenAI tools to engineering, marketing, sales and even event planning, a template for spreading one AI vendor across departments.
What happened. Merge's teams run GPT-6 Astra, ChatGPT and Codex across their daily work, from checking product code to supporting marketing. Staff also lean on AI to organise executive dinners and get sales updates ready.
Why it mattersTeams building AI assistants on AWS can drop in cited web answers with one call instead of building the research steps themselves.
What happened. Amazon Bedrock AgentCore now offers Web Answer, a generally available tool that takes a plain-language question and returns one finished, sourced reply drawn from current web pages. It extends Web Search, which AWS released in June and which only located web content.
Why it mattersIf you run AI agents at scale, the processor handling their actions, not only the AI chip, now affects speed and capacity.
What happened. NVIDIA says its Vera CPU is built to handle the steps AI agents take between reasoning, such as searching, querying databases and running code. The pitch is more agents running at once within a fixed power budget.
Why it mattersTeams letting AI agents write production code get a way to check it against a written spec instead of reading every file.
What happened. Microsoft Research released PDD Kit, a tool that converts a developer's goals into a formal specification, then mathematically proves the code an AI coding agent writes against it. Where proof isn't possible, it falls back to tests, and Microsoft demonstrates it on a full-stack web app.
Why it mattersNon-engineers on your team can turn a rough sketch into a working prototype in one sitting, before anyone books developer time.
What happened. Google's Gemini can turn an uploaded image of a hand-drawn app layout, habit tracker or game board into working code. Users pick the Canvas tool, ask for the app to be built, then preview it and tweak colors, fonts or features as they go.
8 stories, about 3 min 10 sec. Nothing was left out to fill the time.
“I never dreamed about success. I worked for it.”