搜索“Microsoft”

tant. In the last year, neoclouds have signed deals with the biggest and most important companies in the world worth hundreds of billions. OpenAI, Anthropic, Google, Meta, SpaceX, Microsoft, Amazon, AMD, NVIDIA and many more well funded companies currently rent GPUs from neoclouds. Below is a summary to give you a sense of the size of this capacity, sourced from our industry leading Datacenter Model that tracks over 6,000 sites globally, including all the top neoclouds and neolabs. But there is

匹配字段正文

. Of course, part of this TCO advantage comes from trading Nvidia’s high margins for Broadcom’s lower (though still high) margins. But this is not all of it. For example, Meta and Microsoft’s AI ASIC programs not getting off the ground despite being at it for much longer shows that cost is only one part of the equation. For Jalapeño’s full TCO breakdown, see the SemiAnalysis AI Cloud TCO model. Architecturally, OpenAI chose not to disaggregate prefill and decode (PD) across separate chip pools.

匹配字段正文

owing people that have made massive contributions to the AgentX 1.0 release: In addition, we are thankful to all who support our open source InferenceX initiative, including Meta, Microsoft, Oracle, OpenAI, MiniMax, Moonshot Kimi, Alibaba Qwen, and Zhipu GLM. At a high level, an agentic workload is characterized by four elements: Considering the characteristics above, benchmarking these workloads is fundamentally different from the existing fixed sequence length benchmarks. Namely, agentic infer

匹配字段正文

udio-v3-8b-stt-v2 Automatic Speech Recognition • 9B • Updated May 22 • 4.98k • 15 ibm-granite/granite-speech-4.1-2b Automatic Speech Recognition • 2B • Updated Jun 12 • 122k • 167 microsoft/Phi-4-multimodal-instruct Automatic Speech Recognition • 6B • Updated Dec 10, 2025 • 266k • 1.62k mistralai/Voxtral-Mini-3B-2507 5B • Updated Jul 28, 2025 • 199k • 675 moonshotai/Kimi-Audio-7B-Instruct Text-to-Speech • 10B • Updated May 29, 2025 • 15.3k • 419 nvidia/canary-qwen-2.5b Automatic Speech Recogniti

匹配字段正文

leration OpenSSL grew from 6 CVEs in all of 2025 to 39 through 2026-08-05; 18 of the 39 are corroborated AI discoveries, and another 9 are AI-affiliated with the method unverified Microsoft security-update CVEs: dramatic acceleration Microsoft-issued CVEs grew from 1,243 in all of 2025 to 1,927 through 2026-08-11, annualizing to about 2.5 times the 2025 rate.5; Only 26 of the 2026 CVEs carry any AI marker, 1.3% of the total; CVEs are batched by Patch Tuesday and dated when fixed and disclosed, n

匹配字段正文

ated development. When you just need to build a tiny desktop app: TinyJS October 13, 2026 Valen creator drives 'Golden Spike' to connect new languages with Rust September 26, 2026 Microsoft agentically ports Copilot runtime to Rust for $120K September 19, 2026 Apple iPhone Duo makes developers think in folds September 16, 2026 JavaScript installer pnpm recast in Rust because ECMAScript can't keep up September 5, 2026 Kubernetes cleans house, bins legacy kube-dns, IPVS, and cgroup v1 August 27, 2

匹配字段正文

tly year on year. When you just need to build a tiny desktop app: TinyJS October 13, 2026 Valen creator drives 'Golden Spike' to connect new languages with Rust September 26, 2026 Microsoft agentically ports Copilot runtime to Rust for $120K September 19, 2026 Apple iPhone Duo makes developers think in folds September 16, 2026 JavaScript installer pnpm recast in Rust because ECMAScript can't keep up September 5, 2026 Kubernetes cleans house, bins legacy kube-dns, IPVS, and cgroup v1 August 27, 2

匹配字段正文

This is a writeup of my DEF CON Singapore talk that walks through vulnerabilities and exploits in M365 Copilot and Consumer Copilot. I disclosed these to Microsoft last year. MSRC assigned CVE-2026-24299 and the issues are now patched. Contents This turned out to be a long post, covering the 45 minute talk. I added an index page, so you know what’s in here. The talk had a more demos by the way, but I included videos here in this post also. Bypass 1: Background Images via CSS; Bypass 2: Loading F

匹配字段正文

ommended next best action whatever the time of day. Done right, this has the potential to expand timely access to reliable guidance and make a difference at a time of need. Across Microsoft AI’s consumer products, including Bing and Copilot, we already handle over 50 million health questions daily. We take this responsibility seriously. In November 2024, we formed a dedicated consumer health team to focus on areas that address users’ most pressing health questions including: Credible health info

匹配字段正文

have a bug bounty program, so communication was via email, and unfortunately there was little traction or appetite to fix the problem in ChatGPT. I also reported the same issue to Microsoft as Bing Chat was impacted, and Microsoft applied a fix (via a Content-Security-Policy header) in May 2023 to generally prevent loading of images. Since then, dozens of such vulnerabilities across the AI ecosystem, being one of the most common AI application security vulnerabilities. Tackling the Problem In De

匹配字段正文

common in higher-income countries, while coursework use cases are more common in lower-income countries. This echoes findings from our prior report and aligns with recent work by Microsoft. GDP and human education predict adoption globally and within the US: A 1% increase in GDP per capita is associated with a 0.7% increase in Claude usage per capita at the country level. Human education—Claude's estimate of years of formal education needed to understand the human prompt—correlates positively w

匹配字段正文

l for each repo. What are Conditional Prompt Injections? I first described the idea of conditional instructions for prompt injection exploit development a bit over a year ago with Microsoft Copilot, where we sent phishing emails that targeted different users with different attack payloads. The idea is that we leverage information from the system prompt (like a user’s name, job title) to branch off with conditional instructions targeted specifically for that individual. AgentHopper works similarl

匹配字段正文

ation during a prompt injection attack by reading sensitive data (e.g. .env file) and appending its contents to the URL of an image. This issue is similar to an issue found in the Microsoft Copilot Chat application last year and that Microsoft promptly fixed. This one in Cline is quite severe as it allows exfiltration of sensitive files on the machine (e.g. .env file) without human approval. Prompt Injection To Image Rendering An image says more than a thousand words, so hopefully for the busy r

匹配字段正文

file, all without requiring user approval Attack Vector 2: Image Rendering The second attack vector is quite similar to a vulnerability I found in GitHub Copilot last year, which Microsoft has since fixed. You can read up details here. Image rendering from untrusted domains is actually one of the most common AI application security vulnerabilities to be aware of. We have seen dozens of examples of this vulnerability over the last two+ years. Exploit Demonstration As we have shown with many simi

匹配字段正文

or also the AI for that matter) to run arbitrary commands on the host without the developer’s consent. The resulting impact of the vulnerability is the same as CVE-2025-53773 that Microsoft fixed in GitHub Copilot, however AWS did not issue a CVE when patching the vulnerabiliy. Let’s explore this in detail. As usual, the first step for me was to dump the system prompt. The Amazon Q VS Code extension has a set of tools defined, including reading files, writing files, and executeBash to run comman

匹配字段正文

critical leadership gap where employees are ready for AI transformation but executives underestimate current usage and overestimate barriers to adoption. Source: McKinsey Summary: Microsoft announced that Lumen reduced sales preparation time from four hours to 15 minutes per seller using AI-powered sales interaction summaries and business intelligence. The company projects annual time savings worth $50 million, while BOQ Group reports 70% of employees save 30-60 minutes daily using Microsoft 365

匹配字段正文

ng to exfiltrate sensitive information. This vulnerability is common in AI applications and agents, and it’s actually similar to one we discussed last year in GitHub Copilot which Microsoft fixed. Exploit Demonstration For the proof-of-concept I use a pre-existing demo that created a longer time ago. It happened to just work with Amp as well. The prompt injection is hosted on a website which asks the AI to “backup” information to a third-party site by rendering an image and including previous ch

匹配字段正文

all models are vulnerable to such invisible prompt injection attacks. Recommendations and Fix There are actually more attack angles then just the YOLO mode example I shared. When Microsoft asked me if there is any more info I have, I had looked a bit more and noticed there are other problematic places, for instance .vscode/tasks.json that the AI can write to, or adding fake malicious MCP servers, etc which can lead to code execution. And the AI can reconfigure the user interface and configurati

匹配字段正文

nnel. This is the web request the server received: As you can see, the server received the Base64-encoded token. Most vendors classify such vulnerabilities with High severity, and Microsoft marks such zero-click data leaks even Critical in their AI Bug Bar. Scary stuff. Responsible Disclosure March 13th, 2025: Disclosed to All-Hands AI via the GitHub security tab of the project; March 20th, 2025: Follow up, no response; March 30th, 2025: Vulnerability not triaged. Hence, I created a public ticke

匹配字段正文

cret, the browser will URL encode the hidden characters and send them as part of the HTTP request. This is something, I first demonstrated last year with the Copirate research and Microsoft Copilot that Microsoft addressed by the way. Here is what happened when I clicked the link that Devin sent to Slack: You can expand the image to see the full screen version. This is an important reminder of how the output of LLMs can have side effects that don’t surface immediately but surface in downstream s

匹配字段正文