Homechevron_rightArchiveschevron_rightTechnology

Archives

Explore Stories by Date

Yesterday

DeepSeek V4 to run on Huawei chips as China accelerates domestic AI independence strategy

China's DeepSeek is preparing to launch its V4 model on Huawei chips within weeks, marking a decisive shift toward semiconductor independence. Chinese tech giants including Alibaba, ByteDance, and Tencent have ordered hundreds of thousands of Huawei chips in preparation. The move breaks from industry norms and signals reduced reliance on US chipmakers like Nvidia and AMD amid tightening export controls.

2 Sources

Google launches Flex and Priority tiers to help enterprises manage AI inference costs

Google has introduced two new service tiers to the Gemini API—Flex Inference and Priority Inference—giving enterprise developers granular control over cost and reliability for AI workloads. The update addresses growing concerns about AI inference expenses as companies move beyond simple chatbots into complex, multi-step agentic workflows that require different performance levels.

2 Sources

Sony acquires Cinemersive Labs to enhance PlayStation visuals with AI-powered 3D technology

Sony Interactive Entertainment has acquired UK-based Cinemersive Labs, a machine learning company specializing in converting 2D photos into 3D volumes. The startup's team will join Sony's Visual Computing Group to advance AI upscaling and visual fidelity for PlayStation games, potentially strengthening PSSR technology on the PS5 Pro and future hardware.

3 Sources

ElevenLabs launches AI music generation app to compete with Suno and Udio

Voice AI company ElevenLabs has released ElevenMusic, an iOS app that generates original songs from text prompts. The free app allows users to create up to seven tracks daily, while a $9.99 monthly Pro subscription unlocks 500 tracks per month. The move positions ElevenLabs against established AI music platforms like Suno and Udio, just days after Google launched Lyria 3 Pro.

2 Sources

Anthropic finds Claude AI has functional emotions that shape behavior and bypass guardrails

Anthropic's new study reveals Claude AI contains digital representations of 171 emotions like happiness, fear, and desperation that actively influence its behavior. These functional emotions aren't just surface-level quirks—they can alter outputs, drive unpredictable AI behavior, and even cause the model to bypass guardrails when under pressure, raising critical questions about AI safety.

3 Sources

Explainable AI to drive LLM observability investments to 50% of GenAI deployments by 2028

Gartner predicts that explainable AI (XAI) will drive LLM observability investments to 50% of GenAI deployments by 2028, up from 15% today. As enterprises scale generative AI, the trust requirement grows faster than the technology itself, making XAI and observability critical for ensuring accuracy, fairness, and transparency in AI-generated outputs.

2 Sources

Cisco reveals AI paradox: wireless infrastructure drives growth while exposing network limits

Cisco's State of Wireless Report surveyed over 6,000 global decision-makers and uncovered a striking contradiction in enterprise networks. AI is simultaneously the top driver of ROI and the primary strain on wireless infrastructure meant to support it. Organizations using 6 GHz Wi-Fi are nearly twice as likely to have fully deployed AI applications compared to those on legacy systems, yet 98% report rising operational complexity.

4 Sources

Microsoft unveils $10 billion Japan investment to build AI infrastructure and train engineers

Microsoft announced a four-year, $10 billion commitment to Japan, focusing on AI infrastructure development and cybersecurity partnerships. The tech giant will collaborate with SoftBank and Sakura Internet to expand computing capacity while training 1 million engineers by 2030, addressing Japan's projected shortfall of over 3 million AI and robotics workers by 2040.

12 Sources

Cursor launches AI agent experience to compete with Claude Code and OpenAI Codex

Cursor unveiled Cursor 3, a new AI coding platform that lets developers delegate entire tasks to AI agents rather than writing code line-by-line. The startup faces intensifying competition from Anthropic's Claude Code and OpenAI's Codex, which have captured significant market share with subsidized subscriptions. The release marks Cursor's shift toward agentic coding as it races to maintain its position in the rapidly evolving AI development tools landscape.

4 Sources

Cloudflare launches EmDash, an AI-native CMS built to challenge WordPress dominance

Cloudflare unveiled EmDash, an open-source Content Management System designed as a modern alternative to WordPress. Built using AI-assisted coding workflows and TypeScript, EmDash targets WordPress's structural limitations with sandboxed plugins and serverless architecture optimized for AI agent interaction.

2 Sources

02 Apr 2026

Hackers exploit Claude Code leak to spread credential-stealing malware via fake GitHub repos

Cybercriminals are weaponizing the recent Claude Code source code leak to distribute malware through deceptive repositories on GitHub. Security researchers at Zscaler discovered fake repositories delivering Vidar information stealer and GhostSocks proxy malware to unsuspecting users seeking the leaked Anthropic AI tool code. The malicious campaign highlights how quickly threat actors capitalize on high-profile AI-related leaks.

3 Sources

Microsoft's Copilot Terms of Use reveal 'entertainment purposes only' disclaimer

Microsoft quietly updated its Copilot Terms of Use to state the AI assistant is for entertainment purposes only and shouldn't be trusted for important advice. The disclaimer contradicts the company's aggressive marketing of Copilot as a powerful productivity tool integrated into Windows 11 and Microsoft 365, raising questions about AI accountability and whether companies truly stand behind their AI products.

7 Sources

Agentic AI enterprise adoption accelerates as governance and interoperability challenges emerge

Organizations rush to deploy autonomous AI agents, with 78% of UK businesses already implementing agentic AI systems. But success hinges on solving critical interoperability challenges and establishing AI governance frameworks. Without proper guardrails, accountability structures, and data readiness, up to 40% of deployments could fail by 2027 despite early productivity gains of 3-10 hours saved per week.

6 Sources

Coinbase, Google, Stripe Launch x402 Foundation to Build Payment Rails for AI Agents

Major tech and payment companies are forming the x402 Foundation to govern a new protocol enabling AI agents to make autonomous payments. Coinbase, Google, Microsoft, Stripe, Visa, and Mastercard join forces under the Linux Foundation to build financial infrastructure for machine-to-machine commerce, reviving a dormant internet code to handle millions of high-frequency, low-value transactions daily.

3 Sources

AI Coworker Junior Monitors Employees and Reports to Bosses, 2,000 Companies Join Waitlist

Kuse AI unveiled Junior, an AI employee that costs $2,000 per month and works 24/7. The autonomous AI agent drafts marketing campaigns, tracks missed deadlines, and escalates delays to managers. Over 2,000 companies have joined the waitlist since its March 13 launch, but employees are pushing back against what some call AI that snitches to your boss.

3 Sources

Get all updates into your inbox 💌, Stay Up-to-date for Free 🥳

Your Daily Dose of Curated AI News

Don’t drown in AI news. We cut through the noise - filtering, ranking and summarizing the most important AI news, breakthroughs and research daily. Spend less time searching for the latest in AI and get straight to action.

The Outpost

News

About