Homechevron_rightArchiveschevron_rightTechnology

Archives

Explore Stories by Date

Today

Synchrony Partners with OpenAI to Enable Direct Purchases and Rewards in ChatGPT

Synchrony Financial announced an enterprise collaboration with OpenAI to integrate financing and rewards into AI-powered shopping experiences. The credit card issuer for Amazon, Walmart, and Lowe's will allow customers to complete purchases directly inside ChatGPT using their store cards, marking a significant shift toward agentic commerce.

2 Sources

Sainsbury's Pauses AI Facial Recognition at London Store After Customer Wrongly Accused of Shoplifting

Sainsbury's has suspended its Facewatch AI facial recognition system at its East Dulwich store after Matt Arnold was wrongly flagged as a shoplifter and ejected. The incident highlights growing concerns about AI-driven surveillance accuracy and staff reliance on automated alerts without applying human judgment in retail security.

2 Sources

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 AI platform to accelerate drug development with 40% efficiency gains in clinical trials

Tata Consultancy Services introduced TCS ADD AgentHub, an agentic AI platform designed for regulated pharmaceutical environments. The platform enables pharma companies to deploy AI agents at scale across drug development and clinical trials while maintaining regulatory compliance. Early implementations show up to 40% efficiency gains in clinical data management and 30% cost savings in safety case processing.

3 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, outpacing Meta and Google in open-weight race

Alibaba's Qwen family of AI models reached 3 billion downloads in six months, dwarfing Meta's 227 million and Google's 418 million. The Chinese tech giant launched a new laptop-ready model and released weights for its most powerful system, Qwen3.8-Max, cementing its dominance in open-weight AI as US-China competition intensifies.

7 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 Reveals AI Facial Recognition Smart Glasses That Identify People and Create Videos

Meta submitted a patent application for AI-powered smart glasses featuring facial recognition technology that identifies people and automatically generates media files. The 44-page filing details a system that detects individuals, logs their actions, and creates highlight reels from recorded footage. Privacy advocates warn the technology could enable stalking and harassment, calling it an escalation of privacy harms.

4 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

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