AI in News – Curated Headlines That Matter Today

Explore Stories by Date

17 Dec 2025

Browser extensions with 8 million users caught secretly harvesting AI conversations

Popular browser extensions marketed as VPNs and ad blockers are intercepting and selling complete AI chatbot conversations from millions of users. Cybersecurity firm Koi discovered eight extensions, including Urban VPN Proxy with 6 million installs, that capture every prompt and response from platforms like ChatGPT, Claude, and Gemini. The data is transmitted to third-party marketing firms, with no way to disable collection except complete uninstallation.

11 Sources

House advances SPEED Act to fast-track permits for AI data centers and chip manufacturing

The House cleared a critical procedural vote for the SPEED Act, a bipartisan bill backed by OpenAI, Meta, and Microsoft to accelerate federal permits for AI infrastructure. The legislation would shorten environmental review timelines and limit legal challenges to 150 days, but faces internal Republican opposition over an amendment restricting presidential authority on energy permits.

2 Sources

OpenAI Hires George Osborne to Lead Global Stargate Expansion and Government Partnerships

OpenAI has appointed former UK Chancellor George Osborne to head OpenAI for Countries, an initiative to expand its ambitious $500 billion Stargate project beyond US borders. Osborne joins a growing trend of British politicians moving to major tech giants, following Nick Clegg at Meta and Rishi Sunak at Anthropic and Microsoft. The move raises questions about tech companies acting as quasi-governmental actors while OpenAI faces financial concerns over its expansion plans.

12 Sources

DoorDash launches Zesty, an AI-powered app that transforms how you discover new restaurants

DoorDash has launched Zesty, an AI-powered social app designed to simplify restaurant discovery in San Francisco and New York. The app uses an AI chatbot to deliver personalized recommendations by aggregating information from DoorDash, Google Maps, TikTok, and other sources. This marks DoorDash's strategic expansion beyond food delivery into social discovery and local commerce.

5 Sources

Meta AI glasses gain Conversation Focus to amplify voices in noisy environments

Meta rolled out its v21 software update for AI glasses, introducing Conversation Focus to help users hear conversations better in crowded spaces. The update also brings AI-powered Spotify integration that plays music matching your view, plus Telugu and Kannada language support for users in India.

10 Sources

OpenAI rolls back ChatGPT's model router, defaults free users to cheapest AI model

OpenAI has quietly reversed a major change affecting hundreds of millions of ChatGPT users. The company rolled back its model router system for free and Go tier subscribers, now defaulting them to GPT-5.2 Instant instead of automatically routing complex queries to advanced reasoning models. The shift appears to balance cost management with user control, though it raises questions about response quality for sensitive queries.

4 Sources

Nvidia and SK Hynix develop AI SSD targeting 100 million IOPS to solve inference bottlenecks

Nvidia and SK Hynix are jointly developing an AI SSD called Storage Next, promising 100 million IOPS by 2027—ten times faster than current enterprise drives. The collaboration aims to eliminate data storage bottlenecks that throttle AI inference workloads, as the industry shifts focus from training to inference where storage speed becomes critical.

3 Sources

16 Dec 2025

Meta releases SAM Audio AI model to isolate and edit sounds with simple text prompts

Meta has launched SAM Audio, an open-source AI model that simplifies audio editing by isolating specific sounds from complex recordings using text, visual, or time-based prompts. The unified multimodal model can separate voices, instruments, and background noise without manual editing, though it raises questions about privacy safeguards and struggles with similar overlapping sounds.

4 Sources

Red Hat acquires Chatterbox Labs to strengthen AI safety and security guardrails

Red Hat has completed its acquisition of London-based Chatterbox Labs, adding critical AI safety and security capabilities to its portfolio. The deal brings tools to monitor AI models for bias, toxicity, and vulnerabilities as enterprises accelerate AI deployments from experimentation to production at scale.

2 Sources

OpenAI reveals plan to transform ChatGPT into an operating system with new executive hire

OpenAI is moving beyond traditional apps by transforming ChatGPT into an operating system. The company hired Glen Coates as Head of App Platform to lead this evolution, which would make ChatGPT a central software layer for apps and services rather than just another chatbot interface.

2 Sources

Senators probe tech giants as AI data centers drive electricity costs up, communities fight back

Three Democratic senators are investigating whether AI-driven data centers are raising residential electricity bills, citing increases up to 267% in some regions. Meanwhile, grassroots opposition has blocked or delayed $98 billion worth of data center projects in 2025 alone. Tech giants including Amazon, Google, Microsoft, and Meta face scrutiny over confidential utility contracts and broken promises to shield consumers from infrastructure costs.

7 Sources

OpenAI launches GPT Image 1.5 as AI image generator war with Google intensifies

OpenAI released GPT Image 1.5, a new image generation model that generates images up to four times faster than its predecessor and costs about 20 percent less through the API. The release comes as OpenAI responds to Google's viral Nano Banana Pro model, with CEO Sam Altman having declared a 'code red' after Gemini 3's success. The new model promises better instruction-following and photorealistic image manipulation through a dedicated creative studio interface in ChatGPT.

17 Sources

Lovable raises $330M at $6.6B valuation as vibe-coding startup triples worth in five months

Stockholm-based Lovable has raised $330 million in Series B funding at a $6.6 billion valuation, more than tripling its worth since July. The AI coding startup reached $200 million in annual recurring revenue within a year of launch, letting users build apps using text prompts without any coding experience. Major tech investors including CapitalG, Menlo Ventures, and Nvidia backed the round.

11 Sources

Allen Institute for AI releases Molmo 2, challenging Google and OpenAI with open video analysis

The Allen Institute for AI unveiled Molmo 2, a family of open-source AI vision models that can watch, track, and analyze videos with precision. The model surpasses Google's Gemini on video tracking tasks while using just 9 million training videos compared to Meta's 72.5 million. Unlike closed systems from tech giants, Molmo 2 is fully open, releasing model weights, training code, and datasets publicly.

2 Sources

Databricks raises $4B at $134B valuation as AI business drives 55% revenue growth

Databricks has closed a $4 billion Series L funding round at a $134 billion valuation, just three months after hitting $100 billion. The data intelligence company now generates $4.8 billion in annual revenue, with over $1 billion from AI products. The funding will accelerate development of Lakebase, Agent Bricks, and Databricks Apps while supporting acquisitions and global expansion.

6 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