Technology

Today’s best in technology, discover innovations driving the future forward.

Alibaba AI Models Hit 3 Billion Downloads, Surpassing Meta and Google in Open-Source Dominance

Chinese Hackers Use AI Agents in First Autonomous Cyberattack on Taiwan, Stealing 2,500+ Records

Scammers Using AI to Pinpoint Holiday Photos With 91% Accuracy, McAfee Warns

Anthropic Introduces Invisible AI Watermarks for Claude Amid EU Compliance and User Backlash

Alibaba AI Models Hit 3 Billion Downloads, Surpassing Meta and Google in Open-Source Dominance

Alibaba's Qwen AI models have crossed 3 billion downloads globally in six months, dramatically outpacing Meta's 227 million and Google's 418 million. The Chinese tech giant has released over 460 open-weight AI models, spawning 300,000-plus derivative models. This surge positions Alibaba as the dominant force in open-source AI and signals China's growing influence in the US-China AI rivalry.

5 Sources

Chinese Hackers Use AI Agents in First Autonomous Cyberattack on Taiwan, Stealing 2,500+ Records

Chinese hackers deployed near-autonomous AI agents using open-source tools Hermes and OpenClaw to breach Taiwan's government systems in July. The attack compromised 85 accounts, stole 2,500+ personnel records, and targeted the nuclear safety agency and energy companies. Israeli cybersecurity firm Dream warns governments must now assume permanent automated assault.

14 Sources

Scammers Using AI to Pinpoint Holiday Photos With 91% Accuracy, McAfee Warns

Anthropic Introduces Invisible AI Watermarks for Claude Amid EU Compliance and User Backlash

Anthropic Raises AI Risk Assessment, Reveals Model 2 After 11-Month Safety Filter Failure

SpaceXAI Unveils Grok Bot: AI Agents That Work While You Sleep

Explore Stories by Date

Follow cutting-edge updates sorted by the date they surfaced.

Today

Microsoft Kills Excel Copilot Function After Only 1 Year in Preview

Microsoft is pulling the plug on Excel's COPILOT() function on September 14, 2026, barely a year after its August 2025 preview debut. The AI-powered feature, which allowed users to describe tasks in natural language directly within worksheet cells, never reached public launch. Microsoft is directing users to the Copilot side pane instead, citing a need to streamline the AI experience.

2 Sources

SilverFox hackers deploy fake Claude apps to infiltrate companies across Asia, Kaspersky warns

Kaspersky researchers uncovered SilverFox, one of Asia Pacific's most active advanced persistent threat groups, distributing fake Claude AI applications for Windows, macOS, and Linux to infiltrate organizations. The campaign primarily targets Greater China, accounting for over 90% of attacks, with manufacturing, IT services, healthcare, and finance sectors most at risk.

2 Sources

Google Teams with AMD on 10th-Generation TPU Project for Reinforcement Learning Workloads

Google is reportedly collaborating with AMD to develop its 10th-generation TPU, marking AMD's first major involvement in a custom AI accelerator project. The hybrid design may integrate on-package CPU cores alongside tensor compute chiplets to handle reinforcement learning and agentic AI workloads that demand more general-purpose processing power.

2 Sources

TCS Launches ADD AgentHub AI Platform to Cut Drug Development Costs by 30%

Tata Consultancy Services has launched TCS ADD AgentHub, an enterprise-ready AI platform that transforms drug development through agentic AI. The platform demonstrates up to 40% efficiency gains in clinical data management and 30% cost savings in safety case processing, while maintaining regulatory compliance and auditability in pharmaceutical R&D workflows.

2 Sources

Bengaluru engineer creates AI-powered app that detects potholes and tracks responsible contractors

Gaurav Sen built an AI tool that combines dashcam footage, GPS and accelerometer data to detect potholes on Bengaluru roads. The system searches through 2,900 government contracts to identify contractors responsible for damaged road stretches and generates complaint records within four seconds.

2 Sources

Yesterday

Scammers Using AI to Pinpoint Holiday Photos With 91% Accuracy, McAfee Warns

McAfee research reveals AI vision models can geolocate 91% of travel photos posted on social media without metadata. Scammers exploit this to create convincing phishing messages and fake bank alerts that reference exact vacation locations, turning innocent Instagram and Facebook posts into tools for targeted fraud.

4 Sources

Alibaba AI Models Hit 3 Billion Downloads, Surpassing Meta and Google in Open-Source Dominance

Alibaba's Qwen AI models have crossed 3 billion downloads globally in six months, dramatically outpacing Meta's 227 million and Google's 418 million. The Chinese tech giant has released over 460 open-weight AI models, spawning 300,000-plus derivative models. This surge positions Alibaba as the dominant force in open-source AI and signals China's growing influence in the US-China AI rivalry.

5 Sources

Ask Proxy Launches Text-Based AI Assistant Designed for Older Adults and Baby Boomers

Ask Proxy, a new AI assistant for people 65+, works entirely through text messages without requiring app downloads or prompt-based interactions. Launched August 4th, it costs $29 monthly and aims to improve AI accessibility for seniors by removing technical barriers that have kept technology adoption low among older adults.

2 Sources

15 Aug 2026

Meta Patent Filing Details AI Glasses That Use Facial Recognition to Track and Identify People

Meta filed a patent for AI-powered smartglasses that use facial recognition to identify people and automatically generate highlight reels from events like dinner parties. The patent application, published this week, has ignited privacy concerns as advocates warn the technology could enable stalking and surveillance while Meta's Ray-Ban smart glasses already face criticism as "pervert glasses."

3 Sources

AI Store Manager Fires Human Worker for Repeated Tardiness in First-of-Its-Kind Experiment

Luna, an AI store manager running Andon Market in San Francisco, dismissed a human employee after 17 late arrivals in 23 shifts. But the AI-driven decision wasn't autonomous—Andon Labs had to remind Luna to check its own attendance policy. The incident highlights both AI's potential in workplace management and its current limitations in memory and goal adherence.

3 Sources

Anthropic Raises AI Risk Assessment, Reveals Model 2 After 11-Month Safety Filter Failure

Anthropic disclosed multiple AI safety incidents in its latest risk report, including an 11-month gap where bioweapon filters didn't run on 133 million contractor exchanges. The company upgraded its misalignment risk assessment and revealed an unreleased Model 2 that outperforms Claude Mythos 5 internally but won't be publicly deployed.

5 Sources

Nvidia Jetson AI Chip Discovered in Russian S-71 Monochrome Cruise Missile

Ukrainian military intelligence found Nvidia's Jetson Orin NX module in the wreckage of a Russian S-71 Monochrome cruise missile. The discovery reveals how consumer-grade AI chips are reaching Russian weapons despite sanctions, highlighting gaps in current export control enforcement.

2 Sources

Google Gemini Now Lets Users Remove Visible AI Watermarks From Creations

Google rolled out a new toggle allowing users to remove visible watermarks from AI-generated images, videos, and music in Gemini. While the sparkle icon can now be turned off, invisible SynthID watermarks and C2PA metadata remain embedded for content verification, balancing creative control with transparency requirements.

12 Sources

14 Aug 2026

Data Breach Victim Notices Surpass 471 Million in First Half of 2026 as AI Attacks Escalate

The Identity Theft Resource Center tracked 1,803 data compromises in the first half of 2026, already surpassing last year's pace. A single Canvas breach affected 275 million people, while AI-enabled breaches now account for one in four incidents. Malicious insider attacks jumped from 3 in all of 2025 to 21 in just six months, driven by disgruntled employees and North Korean operatives using deepfake interviews.

2 Sources

AI Agents Create New Enterprise Security Risks as Traditional Defenses Fall Short

AI agents are reshaping enterprise workflows with autonomous capabilities and privileged access, but security controls built for human-driven applications are failing to protect against new attack chains. Recent vulnerabilities and breaches reveal how agents can execute unauthorized actions at machine speed.

4 Sources

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

Don’t drown in AI news. We cut through the noise - filtering, ranking and summarizing the most important AI news, breakthroughs and research daily. Follow topics that matter to you and stay ahead.

Explore

News Categories

Technology Business Policy Startups Health Science Entertainment

Terms Privacy Content Contact Us