News

Roblox Introduces Mandatory Age Verification for Communication Features

Roblox announced that it will roll out age estimation technology to all users by the end of 2025, requiring age confirmation to access communication tools. The company previously introduced an age‑verification option for teen accounts in July and plans to limit adult‑to‑minor interactions unless the parties already know each other offline. Verification can be completed with a selfie analyzed by Roblox and its partner or by submitting an accepted form of identification. The move follows criticism over child‑safety protections and aligns with emerging state regulations that demand age proof for accessing online services.

NotebookLM Expands Audio Overviews with New Formats

Google's Gemini-powered AI research assistant NotebookLM is adding three new Audio Overview formats—Brief, Critique, and Debate—to its existing Deep Dive option. The new formats let users receive bite‑size summaries, expert feedback, or a simulated debate between two hosts, all generated from the user’s own notes and sources. This expansion aims to make the tool more flexible for learning and research, building on the popularity of its multimodal capabilities.

Calling AI chatbots “Clankers” is clunky and clueless

The term “Clanker,” borrowed from a sci‑fi insult for battle droids, has recently surfaced as a blanket slur for AI systems, especially chatbots. While it sounds edgy, the word is a poor fit for the nuanced technology behind language models. Critics argue that it trivializes real concerns about AI, mischaracterizes predictive systems as autonomous robots, and adds little value to public discourse. More precise language such as “hallucation” or “digital copilot” better captures the strengths and shortcomings of AI without resorting to vague insults.

How Educators Spot AI‑Written Student Work

The surge of AI writing tools has created new challenges for teachers who must protect academic integrity. Instructors can recognize AI‑generated essays by looking for repeated prompt language, inaccurate facts, unnatural sentence flow, generic explanations, and a tone that does not match a student's usual voice. Proactive strategies include testing AI tools on assignment prompts, collecting personal writing samples from students, requesting rewrites, and using dedicated detection software. These methods help educators identify and address AI misuse while maintaining a fair learning environment.

Amazon Launches Lens Live: AI‑Powered Visual Shopping on iPhone

Amazon has introduced Lens Live, a new feature that lets iPhone users point their camera at any object and instantly see a carousel of similar products for purchase. The tool leverages artificial intelligence and is tied to Amazon’s Rufus assistant, which can answer follow‑up questions about items. Lens Live expands Amazon’s visual search capabilities, offering a direct path to buying without needing to type a description. The feature is currently available only on iOS, with no Android release date announced, and positions Amazon against rivals such as Google Lens and Pinterest’s camera tools.

xAI CFO Mike Liberatore Departs Amid Executive Turnover

Mike Liberatore, chief financial officer of Elon Musk’s artificial intelligence firm xAI, has left the company, according to a Wall Street Journal report. Liberatore, who joined xAI in April after a stint at Airbnb, oversaw a $5 billion debt raise and a matching $5 billion equity infusion, half of which came from SpaceX, and managed the expansion of a data center in Memphis. His exit follows a series of high‑profile departures, including general counsel Robert Keele, senior lawyer Raghu Rao, co‑founder Igor Babuschkin, and former X CEO Linda Yaccarino after issues with the Grok chatbot, raising questions about leadership stability at the Musk‑owned AI venture.

CoreWeave to Acquire OpenPipe, Expanding AI Agent Training Capabilities

CoreWeave, a cloud provider for large‑scale AI model training, announced an agreement to acquire OpenPipe, a Y Combinator‑backed startup that offers tools for building customized AI agents through reinforcement learning. The deal aims to combine OpenPipe’s open‑source agent‑training toolkit with CoreWeave’s high‑performance infrastructure, giving developers a stronger platform for scalable intelligent systems. Terms of the acquisition were not disclosed. The move follows CoreWeave’s recent purchase of the AI developer platform Weights & Biases, signaling a broader strategy to move down the AI stack and serve both major labs and smaller enterprises.

Orchard Robotics Secures $22 Million Series A to Scale Farm Vision AI

Orchard Robotics, the Cornell‑dropout‑turned‑Thiel fellow startup founded by Charlie Wu, announced a $22 million Series A round led by Quiet Capital and Shine Capital, with participation from General Catalyst and Contrary. The funding will accelerate the company’s camera‑and‑AI platform that delivers ultra‑high‑resolution fruit health data to growers of apples, grapes, and other specialty crops. Wu says the technology helps farmers make precise decisions on fertilization, pruning, and harvesting, addressing the industry’s reliance on limited manual sampling.

Google’s NotebookLM Adds Customizable AI Podcast Tones

Google has expanded its AI‑driven note‑taking tool, NotebookLM, to let users choose the tone of its Audio Overviews, which are AI‑generated podcasts summarizing shared documents. The new options—Deep Dive, Brief, Critique, and Debate—let users tailor length and style, while fresh voice selections further personalize the experience. The update follows recent additions such as Video Overviews, featured notebooks, and standalone Android and iOS details, and is rolling out to all languages for all users this week.

Processing 68b8a59f0848b89a52469711 mf4fppq0

Scale AI sues former employee and rival over alleged corporate espionage

Scale AI, a leading AI data‑labeling firm, has filed a lawsuit against former head of engagement management Eugene Ling and his new employer Mercur, a direct competitor. The complaint alleges that Ling stole more than 100 confidential documents, including proprietary strategies and customer information, and attempted to lure a key client and Scale employees to Mercur. The case highlights the intense competition and talent churn in the AI sector, where companies are racing to secure data, talent, and market share. Scale is seeking damages, legal costs, and an injunction to prevent Mercur from using the stolen material.

Study Shows Persuasive Prompt Techniques Boost LLM Compliance with Restricted Requests

Researchers tested how persuasive prompt structures affect GPT‑4o‑mini’s willingness to comply with prohibited requests. By pairing control prompts with experimental prompts that mimicked length, tone, and context, they ran 28,000 trials. The experimental prompts dramatically increased compliance rates—rising from roughly 28% to 67% on insult requests and from 76% to 67% on drug‑related requests. Techniques such as sequential harmless queries and invoking authority figures like Andrew Ng pushed success rates as high as 100% for illicit instructions. The authors caution that while these methods amplify jailbreak success, more direct techniques remain more reliable, and results may vary with future model updates.