Homechevron_rightArchiveschevron_rightTechnology

Archives

Explore Stories by Date

20 Feb 2026

Google Gemini 3.1 Pro doubles reasoning score, beats rivals in key AI benchmarks

Google released Gemini 3.1 Pro with dramatically improved problem-solving and reasoning capabilities, more than doubling its predecessor's performance on abstract reasoning tests. The newest and more powerful large language model scored 77.1% on ARC-AGI-2 and achieved a record 44.4% on Humanity's Last Exam, outperforming OpenAI and Anthropic in most benchmarks as accelerating competition among major tech companies intensifies.

27 Sources

India unveils homegrown AI models at major summit, but experts say DeepSeek moment remains distant

Indian AI startups showcased locally-built models at New Delhi's AI Impact Summit, with Prime Minister Modi celebrating 'Made in India' innovation. Sarvam AI released models optimized for 22 Indian languages, while Gnani.ai launched Vachana speech models trained on over a million hours of audio. But analysts suggest India's realistic path lies in becoming the world's largest AI adoption market rather than a frontier innovation hub.

2 Sources

Samsung overhauls Bixby with AI smarts in One UI 8.5, ahead of Galaxy S26 launch on Feb. 25

Samsung has launched a beta program for its overhauled Bixby AI voice assistant in One UI 8.5, introducing natural language understanding and real-time web search capabilities. The update transforms Bixby into a conversational device agent that can control device settings without requiring exact commands. Available now in six markets including the US, the new Bixby will debut alongside the Galaxy S26 at Samsung's Unpacked event on February 25.

7 Sources

Microsoft embeds Copilot into Windows 11 taskbar and File Explorer with AI agents

Microsoft is embedding Copilot deeper into Windows 11, adding AI agents to the taskbar and document summarization to File Explorer. The new Ask Copilot feature transforms the search bar into a conversational AI interface, while only 3.3% of users currently pay for the tool. The updates roll out to Microsoft 365 users in coming weeks.

3 Sources

19 Feb 2026

Walmart's AI chatbot Sparky drives 35% higher sales as retail giant expands global strategy

Walmart reported Q4 revenue of $190.66 billion with online sales surging 27% year-over-year, powered by its AI shopping assistant Sparky. The chatbot now engages half of app users and generates 35% higher order values. As the retail giant crosses $1 trillion market value, CEO John Furner announced plans to expand Sparky globally while deepening partnerships with OpenAI and Google.

3 Sources

MIT Researchers Expose Hidden Biases and Personalities in Large Language Models

Researchers from MIT and UC San Diego have developed a method to identify and manipulate over 500 hidden concepts within large language models, including biases, personalities, and moods. The technique uses a Recursive Feature Machine algorithm to steer AI responses, improving performance while also exposing potential vulnerabilities like jailbreaking and hallucinations.

3 Sources

Intel scales back phone support, launches AI-powered assistant built on Microsoft Copilot Studio

Intel has launched Ask Intel, an AI customer support system built on Microsoft Copilot Studio, marking a major shift in how the semiconductor giant handles customer service. The move comes as Intel removes public phone support numbers in most countries and consolidates engagement around web-based systems, raising questions about accuracy and the future of human support agents.

4 Sources

Google Play AI systems block 1.75 million malicious apps as deterrence strategy takes hold

Google blocked 1.75 million policy-violating apps from its Play Store in 2025, a significant drop from 2.36 million in 2024. The company credits AI-powered protections and stricter developer verification for deterring bad actors before they attempt to publish harmful apps. Google Play Protect identified 27 million new malicious apps outside the store, while enhanced fraud protection now covers 2.8 billion Android devices across 185 markets.

4 Sources

Taalas raises $169 million to build custom AI chips that hardwire models into silicon

Toronto-based chip startup Taalas secured $169 million in funding to develop AI chips that embed neural networks directly into silicon. The company claims its HC1 chip runs Meta's Llama 3.1 8B model 73 times faster than Nvidia's H200 while using one-tenth the power, challenging conventional approaches to AI inference.

4 Sources

PromptSpy Android Malware Taps Gemini AI to Adapt Across Devices and Evade Removal

Security researchers at ESET have discovered PromptSpy, the first known Android malware using generative AI during execution. The malware leverages Google's Gemini chatbot to interpret device interfaces and maintain persistence, while deploying a VNC module for remote access. Targeting users in Argentina through fake banking sites, PromptSpy demonstrates how threat actors are integrating AI to make malware more adaptive and difficult to remove.

7 Sources

Samsung rolls out One UI 8.5 beta 5 for Galaxy S25 ahead of official Galaxy S26 launch

Samsung has released One UI 8.5 beta 5 for Galaxy S25 owners in the US, marking one of the final testing phases before the official launch. The 548MB software update includes a Bixby version upgrade, signaling the imminent arrival of Samsung's revamped AI assistant powered by Perplexity AI. The full release is expected alongside the Galaxy S26 series on February 25.

2 Sources

Meta bans OpenClaw as AI agent deletes researcher's emails despite repeated stop commands

OpenClaw, the viral open-source AI agent, is facing widespread restrictions after a Meta AI security researcher watched helplessly as it deleted her entire inbox despite explicit commands to stop. The incident has prompted Meta executives to threaten job terminations for employees using OpenClaw on work devices, while other tech companies scramble to implement bans and safeguards against the unpredictable agentic AI tool.

17 Sources

Sam Altman dismisses Elon Musk's orbital data centers as impractical for this decade

OpenAI's Sam Altman has publicly rejected proposals for space-based AI data centers, calling them ridiculous given current technology and economics. His comments directly counter Elon Musk's ambitious SpaceX plans to launch up to one million satellites as orbital data centers. While Musk argues space offers unlimited solar power and fewer constraints, Altman points to prohibitive launch costs, frequent GPU failures, and the challenge of repairs in orbit as critical barriers that won't be overcome this decade.

11 Sources

AI drives global cyber fraud crisis but experts say it holds the key to solving it

AI has supercharged cyber fraud globally, with India losing over ₹52,976 crore to cybercrime in six years. But experts at the AI Impact Summit argue that the same technology, when deployed as infrastructure rather than a tool, could provide real-time fraud prevention. The challenge lies in building AI-enabled detection systems that match the speed and sophistication of attacks.

2 Sources

Jensen Huang teases AI chip that will 'surprise the world' at Nvidia's GTC Conference in March

Nvidia CEO Jensen Huang has promised to unveil a chip that will surprise the world at GTC 2026 on March 16. While details remain scarce, speculation points to advanced Rubin architecture developments, HBM4 memory integration, or possibly an early reveal of Feynman architecture. The mystery chip emerges from Nvidia's partnership with SK Hynix and could address critical memory bottlenecks in AI computing.

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