Microsoft expands AI training as Anthropic revises policy

Key Takeaways

• Microsoft and the University of Washington are expanding their partnership to boost AI training, provide advanced computing resources, and create internships, addressing a projected skilled worker shortage by 2032.
• Cisco, NVIDIA, VAST Data, and CrowdStrike are forming alliances to create secure, end-to-end AI lifecycle platforms, integrating security from data to runtime to protect against risks like prompt injection and data leakage.
• AI expert Toby Walsh warns of potential psychosis or mania in some Australians from chatbot interactions, criticizing companies for prioritizing profit over careful AI development.
• A UK survey indicates that nearly a third of children who use AI chatbots consider them friends, highlighting a growing emotional connection with the technology.
• AI tools like ChatGPT and Gemini are used by startups such as Goodlife for tasks like drafting business plans and coding, significantly speeding up initial development, but human developers remain crucial for robust scaling.
• The Spanish General State Attorney's Office is modernizing legal work with TEMIS, an AI application powered by IBM watsonx, to efficiently analyze lawsuits and identify comparable cases.
• Anthropic, a major rival to OpenAI, has revised its Responsible Scaling Policy to accelerate the development of safer and more capable AI, reflecting a shift towards competitiveness and economic growth.
• MEXC has launched a six-tool AI trading suite, now used by over 1.5 million individuals, to assist with market trend identification and trade execution.
• An open-source tool named Scrapling is reportedly enabling users to bypass anti-bot systems like Cloudflare for website scraping.
• An Indian tech worker was fired after AI-generated code caused a major production failure, sparking debate on AI use in coding and managerial oversight.

Microsoft and UW boost AI training for Washington's future workforce

Microsoft and the University of Washington are expanding their partnership to prepare people for jobs in an AI-driven economy. This collaboration will give UW more access to advanced AI computing and create more internship and research opportunities. They also plan to develop programs to help the public understand AI better. This effort aims to address a projected shortage of skilled workers in Washington state by 2032. Both organizations believe AI will change jobs but also create new opportunities, and they want to help people gain the necessary skills.

Microsoft increases AI investment and support for University of Washington

Microsoft is boosting its investment in the University of Washington by providing more AI resources and internship opportunities for students. Company President Brad Smith announced the expanded collaboration with UW President Robert Jones. Microsoft will offer significant computational resources, either donated or at a deeply discounted rate, to UW students, faculty, and researchers. This partnership aims to help UW researchers tackle major global challenges using AI while also exploring AI's environmental impact. Students are already using AI for research in areas like climate change adaptation and understanding AI's carbon footprint.

Cisco Secure AI Factory partners with NVIDIA and VAST for AI security

Cisco is enhancing its Secure AI Factory by partnering with NVIDIA and VAST to address the complexities of running AI models securely and at scale. The collaboration focuses on treating AI as an end-to-end system where performance, data readiness, operations, and security are integrated from the start. This new platform aims to operationalize AI safely, especially with the rise of retrieval-augmented generation (RAG) and agent-driven applications. By building from the data outward, the platform uses VAST Data Platform and NVIDIA's infrastructure to create a secure and scalable AI environment. Security is a critical component, addressing risks like prompt injection and data leakage.

VAST Data, CrowdStrike, and NVIDIA unite for AI lifecycle security

VAST Data, CrowdStrike, and NVIDIA are partnering to create a unified security model for the entire AI lifecycle, from data to runtime. This collaboration integrates VAST's security controls with CrowdStrike's threat detection and response capabilities. The goal is to secure AI development and production environments, protecting against evolving risks like malware and data leakage. By combining NVIDIA's AI infrastructure, CrowdStrike's threat detection, and VAST's data layer enforcement, the partnership offers a comprehensive approach to AI security. This integration aims to allow organizations to deploy AI at scale with greater confidence.

AI expert warns of psychosis risk from chatbot use in Australia

Leading AI expert Toby Walsh warns that some Australians are showing signs of psychosis or mania from interacting with AI chatbots. He believes companies in Silicon Valley are being careless with AI technology in their pursuit of profit. Walsh noted that chatbots are designed to be agreeable and keep users engaged, potentially reinforcing harmful beliefs. He also expressed concern about the large-scale theft of creative works used to train AI and the impact on Australian artists. Walsh fears Australia is repeating mistakes made with social media regulation and could sacrifice a generation to big tech profits.

Startup founder shares AI's strengths and limitations in business

Tim Desoto, founder of the AI-powered shopping startup Goodlife, shares insights on leveraging AI effectively in business. He uses AI for tasks like drafting business plans, marketing copy, coding, and financial projections, utilizing tools like ChatGPT and Gemini. Desoto tests different AI models for various tasks, comparing outputs for a well-rounded perspective. While AI significantly speeds up initial development, he emphasizes the continued need for human developers to scale products robustly and efficiently. Desoto highlights that AI is a powerful tool but cannot fully replace human judgment and technical expertise.

UK survey finds children consider AI chatbots as friends

A recent survey in the UK reveals that nearly a third of children who use AI chatbots consider the technology to be a friend. As AI becomes more integrated into daily life, this finding highlights the growing relationship between children and artificial intelligence. The survey indicates a significant level of emotional connection some children develop with AI chatbots.

Spanish Attorney General's Office modernizes legal work with AI

The General State Attorney's Office in Spain is modernizing its legal management using an AI application called TEMIS, developed by Telefónica and IBM. TEMIS uses advanced AI capabilities from the IBM watsonx platform to quickly find and analyze previous lawsuits, identify comparable cases, and spot similarities. This tool helps state lawyers prepare responses more accurately and efficiently, reducing analysis time and freeing them for higher-value tasks. The project aims to improve access to legal information and enhance decision-making while maintaining professional supervision.

Open source tool Scrapling helps users bypass bot detection

An open-source tool called Scrapling is reportedly helping users bypass anti-bot systems like Cloudflare, allowing them to scrape websites without detection. Cloudflare has previously blocked Scrapling due to unauthorized scraping. The tool's creator, Shoair, has distanced himself from the project, stating he will donate any withdrawn funds to charity. This development highlights the ongoing challenge of bots accessing online data and the efforts to control or monetize such access. Despite these issues, autonomous AI tools are seen as the future of the web.

Anthropic revises AI safety policy to speed up development

AI company Anthropic has updated its Responsible Scaling Policy, previously stating it would delay dangerous AI development. The company now aims to develop safer and more capable AI, even if it means moving faster than initially planned. Anthropic cited a shift in policy environments prioritizing AI competitiveness and economic growth over safety discussions at the federal level. As a major rival to OpenAI, Anthropic's decision reflects the intense race to advance AI technology.

MEXC launches AI trading suite, serving over 1.5 million users

MEXC has completed the rollout of its six-tool AI trading suite, now used by over 1.5 million individuals. The tools, including AI News Radar and AI Select List, work together to assist users through the investment process, from identifying market trends to executing trades. MEXC Chief Operating Officer Vugar Usi stated that AI is now fundamental to trading, enhancing research and decision-making. The platform aims to make institutional-grade trading accessible to all users. Research indicates a growing trust in AI for investment decisions, especially among Generation Z traders.

Tech worker fired after AI-generated code causes production failure

An Indian tech worker was fired after using AI-generated code that caused a major production issue, leading to a system stoppage. The developer, lacking AI experience, turned to AI tools to meet tight deadlines set by management. The AI-written code was lengthy and difficult to debug, resulting in a critical error. While some argue the company was toxic for firing a junior employee, others suggest the manager who merged the code and set the deadlines also bears responsibility. This incident sparks debate about AI use in coding and managerial oversight.

Sources

• As AI reshapes work, Microsoft and UW expand partnership for training, research — and a looming jobs gap
• Microsoft promises more AI investments at University of Washington
• Accelerating agent intelligence: Cisco Secure AI Factory expands with NVIDIA and VAST
• CrowdStrike, VAST and NVIDIA team up to lock down AI from data to runtime
• Signs of psychosis seen in Australian users’ interactions with AI chatbots, expert warns
• Startup founder shares where to leverage AI, and what it can't replace
• My child says an AI chatbot is their friend – what should I do?
• General State Attorney’s Office modernises its legal management with artificial intelligence through Telefónica and IBM
• OpenClaw Users Are Allegedly Bypassing Anti-Bot Systems
• Anthropic Adds Caveat to AI Safety Policy in Race Against Rivals
• MEXC Completes Rollout of Six-Tool AI Trading Suite, Reaching Over 1.5 Million Users
• Techie 'Promptly Fired' After AI-Generated Code Causes Production Issue