Homechevron_rightArchiveschevron_rightTechnology

Archives

Explore Stories by Date

16 Dec 2025

Klarna launches open standard to help AI agents discover and compare 100 million products

Klarna unveiled its Agentic Product Protocol, giving AI agents access to a live feed of over 100 million products and 400 million prices across 12 markets. The open standard allows merchants to make their inventory discoverable to any AI platform without ads or intermediaries, establishing a foundation for the next generation of agentic commerce.

2 Sources

Google tests faster AI Mode history access on Android, matching iOS style history view

Google is testing a streamlined way to access AI Mode history in its Android app, potentially replacing the Labs shortcut with direct access from the home screen. The update, spotted in version 16.50.55, mirrors an iOS implementation already live, organizing past interactions by date for easier review.

2 Sources

15 Dec 2025

Google Search adds plus button for instant AI Mode file uploads on desktop

Google has quietly introduced a plus button to its search homepage, allowing desktop users to upload images and documents directly for AI-powered analysis through AI Mode. The feature, which appears on the left side of the search bar, represents another step in Google's pivot to AI as competition with OpenAI intensifies.

5 Sources

AI at Work Doubles to 45% as Gallup Survey Exposes Major Communication Gap Between Employers

A new Gallup survey reveals that 45% of U.S. workers now use AI at work at least a few times a year, more than doubling from 21% in 2023. But nearly one-quarter of employees don't know if their employers have adopted AI at all, exposing a stark communication gap. The data shows AI adoption varies dramatically by industry, with tech workers leading at 76% while retail lags at 33%.

5 Sources

Zoom AI Companion 3.0 launches with browser access and free tier for meeting summaries

Zoom unveiled AI Companion 3.0, bringing its AI assistant to the web and opening access to free users for the first time. Basic plan users can now use the assistant for three meetings monthly, including meeting summaries, AI note-taking, and 20 questions. The update introduces agentic automation, document creation tools, and integration with third-party services like Google Drive and Microsoft OneDrive.

4 Sources

Nvidia acquires SchedMD to strengthen open-source Slurm and unveils Nemotron 3 AI models

Nvidia acquired SchedMD, the developer behind Slurm workload management system, while launching the Nemotron 3 family of open AI models. The chipmaker pledges to maintain Slurm as vendor-neutral software while accelerating development for high-performance computing and AI. The move signals Nvidia's push to control more of the AI infrastructure stack while positioning itself as an open-source champion.

11 Sources

CrowdStrike launches Falcon AIDR to secure AI prompts as company calls them 'the new malware'

CrowdStrike unveiled Falcon AI Detection and Response (AIDR), expanding its platform to 32 modules to address the rapidly growing AI attack surface. The company's president Mike Sentonas declares that prompts are the new malware, positioning the solution to protect organizations against prompt injection and malicious AI agent interactions with real-time detection and response capabilities.

2 Sources

Sergey Brin calls retirement 'the worst decision' and returns to Google to lead Gemini AI push

Google co-founder Sergey Brin admits his 2019 retirement was a mistake, revealing he was 'spiraling' before returning to work on Gemini AI. Speaking at Stanford University, Brin acknowledged Google 'messed up' by underinvesting in large language models despite pioneering the Transformer research that powers ChatGPT.

3 Sources

Nvidia launches Nemotron 3 open source AI models as Meta steps back from transparency

Nvidia released its Nemotron 3 family of models in three sizes—Nano (30B), Super (100B), and Ultra (500B parameters)—with full training data transparency. The move positions the chip giant as a major force in open source AI just as Meta signals a retreat from openness, leaving Chinese models to dominate the space.

16 Sources

LG TVs forced Microsoft Copilot installation sparks backlash over unremovable AI shortcuts

LG smart TV owners discovered Microsoft Copilot automatically installed on their devices after a recent webOS update, with no option to remove it. The company has now promised to let users delete the shortcut following widespread criticism, but the incident highlights growing concerns about AI features being forcibly integrated into consumer devices without explicit consent.

17 Sources

GPT-5.2 still can't solve the infamous strawberry question despite billions in AI investment

OpenAI's GPT-5.2, released in December 2025, continues to incorrectly count two r's in the word strawberry when there are actually three. The error stems from the model's tokenization process, which splits the word into st-raw-berry tokens. While other AI models like Claude, Gemini, and Perplexity answer correctly, ChatGPT's tokenized input/output design creates persistent limitations in basic letter counting tasks that a seven-year-old could solve.

2 Sources

AI-Powered Attacks Exploit Identity Security as Threat Actors Automate Cyber Espionage

Threat actors are weaponizing generative AI to automate 80-90% of cyberattacks, targeting identity security through deepfake-enabled remote IT worker scams and sophisticated phishing campaigns. With 75% of organizations experiencing SaaS-related incidents involving compromised credentials, the exploitation of stolen credentials has become the primary attack surface, bypassing traditional security measures entirely.

2 Sources

Tesla starts testing robotaxis in Austin with no safety drivers as Musk pushes autonomous future

Tesla has begun testing its robotaxis in Austin without human safety monitors onboard, marking a significant step toward launching a commercial driverless ride-hailing service. Elon Musk confirmed the development over the weekend after videos of empty Model Y vehicles surfaced online. The company's test fleet has been involved in at least seven crashes since June.

8 Sources

Extremist groups use AI voice cloning and deepfakes to supercharge propaganda and recruitment

Militant organizations from neo-Nazis to the Islamic State are deploying AI voice cloning and generative AI tools to recreate historical speeches and translate propaganda into multiple languages. National security experts warn these technologies enable even poorly resourced groups to produce sophisticated content at scale, helping them recruit new members and expand their reach across social media platforms.

6 Sources

Amazon Kindle AI feature launches with no opt-out for authors as Ask this Book goes live

Amazon rolled out Ask this Book, a Kindle AI feature that provides spoiler-free answers to reader questions about plot and characters. Available now on iOS for thousands of English titles, the in-book AI assistant has sparked controversy as authors and publishers cannot opt out, raising questions about AI copyright and derivative works in publishing.

13 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