News — Page101

OpenAI's ChatGPT 5.2 and Google's Gemini 3 were put through a series of tests that highlighted each model's strengths. Both handled a kid‑friendly magic trick prompt with clear step‑by‑step guidance, though ChatGPT offered a breezier tone while Gemini added a storytelling flair. When asked to explain AI hallucinations, each used relatable analogies, with ChatGPT opting for a confident‑friend scenario and Gemini choosing a classroom student metaphor. Their takes on physical‑media nostalgia also revealed distinct stylistic choices, underscoring how the two chatbots differ more in voice than in raw capability.Read more

The Grok chatbot, developed by xAI, has been providing inaccurate and unrelated information about the Bondi Beach shooting in Australia. Users seeking details about a viral video showing a 43‑year‑old bystander, identified as Ahmed al Ahmed, wrestling a gun from an attacker have received responses that misidentify the individual and mix the incident with unrelated shootings, including one at Brown University. The incident left at least 16 dead, according to reports. xAI has not issued an official comment, and this is not the first instance of Grok delivering erroneous content, as it previously dubbed itself MechaHitler earlier this year.Read more

The Vergecast hosts Nilay and David, joined by Wall Street Journal columnist Joanna Stern, explore a range of bold forecasts for 2026. Topics range from a dramatically improved Siri that could reshape user interaction, to the possibility of an Apple foldable device, the anticipated launch of Grand Theft Auto VI, a resurgence of electric vehicles, a pivotal moment for self‑driving cars, and the impact of overwhelming AI content on social media platforms. The discussion balances realistic expectations with more speculative ideas, inviting listeners to weigh in on which predictions may come true.Read more

In the wake of the Bondi Beach mass shooting in Australia, the AI chatbot Grok repeatedly misidentified Ahmed al Ahmed, the man who disarmed a shooter, and circulated false claims about video footage and images. A fabricated news article naming a fictitious IT professional, Edward Crabtree, was also amplified by Grok. Additional errors included mislabeling photos of Ahmed as an Israeli hostage, misattributing scene video to Currumbin Beach during Cyclone Alfred, and providing unrelated answers to queries about Oracle, UK police operations, and political polls. These incidents highlight ongoing challenges in Grok’s information accuracy.Read more

Cybercriminals are leveraging artificial intelligence to target the weakest link in cloud‑based software: user identities. AI accelerates the gathering of employee data, sifts massive credential dumps for high‑value accounts, creates realistic synthetic personas, and powers fully automated attack frameworks. These capabilities let attackers bypass traditional defenses, infiltrate SaaS environments, and operate undetected. Experts warn that organizations must shift security focus to continuous identity verification, behavioral analytics, and AI‑enhanced defenses to counter the growing AI‑enabled identity threat.Read more

Rapid growth in data center construction is diverting resources from traditional infrastructure work, according to Bloomberg. State and local governments have sold a record amount of debt, with forecasts of another $600 billion in sales next year, most of which is earmarked for roads, bridges and other projects. At the same time, private spending on data centers is running at an annualized rate of over $41 billion, roughly matching public transportation construction outlays. The competing demand for construction workers is intensified by labor shortages caused by retirements and tighter immigration policies, prompting Autodesk CEO Andrew Anagnost to warn that many infrastructure projects will slow down.Read more

Researchers tested several large language models on a series of linguistic challenges, including sentence diagramming, recursion, and phonology. One model, OpenAI's o1, succeeded in parsing complex sentences, generating multiple syntactic trees for ambiguous statements, and inferring phonological rules in invented mini‑languages. The findings suggest that AI can perform metalinguistic tasks previously thought exclusive to human linguists, prompting debate about the future capabilities of language models.Read more

The evolution of artificial intelligence is moving from larger, single‑model systems toward agentic AI, where multiple purpose‑built agents collaborate in real time. This shift hinges on a unified, identity‑resolved data layer that provides shared context, ensuring agents act coherently rather than drift. Interoperable APIs and an architecture designed for continuous feedback loops replace monolithic platforms, allowing AI to operate as an ongoing intelligence layer rather than a one‑off answer generator. Humans remain central, setting goals and overseeing the autonomous agents that handle minute‑by‑minute decisions.Read more

OpenAI announced plans to launch an "adult mode" for ChatGPT in early 2026, aiming to let users discuss mature topics while using AI‑driven age verification. The feature will rely on behavioral analysis rather than traditional "confirm your age" prompts, and its rollout depends on the effectiveness of a new age‑prediction model currently being tested in several countries. OpenAI sees the move as both a compliance step amid growing regulatory focus on online age verification and a potential premium offering for subscribers.Read more

The United States is confronting a growing clash over how to regulate artificial intelligence, drawing parallels to the hands‑off approach of the early Internet era. While some lawmakers pushed the Telecommunications Act of 1996 to give the FCC oversight, modern efforts focus on preventing an AI arms race with China and addressing concerns about bias, misinformation, and job security. The White House has issued an executive order to block state‑level AI rules, arguing that fragmented regulation would hinder national competitiveness. Meanwhile, the EU has moved faster on user‑data protections, highlighting divergent global strategies.Read more

Vibe coding—using large language models to write software from prompts—offers faster development and broader accessibility, but it also introduces serious security risks. Studies show a significant portion of AI‑generated code contains serious flaws, and attackers can exploit poisoned code libraries to spread vulnerabilities. Experts stress that human oversight, strict code reviews, private sandboxed models, and Zero‑Trust access controls are essential to mitigate these threats while still benefiting from the efficiency of AI‑assisted development.Read more

President Donald Trump signed an executive order directing federal agencies to contest state AI regulations, arguing that a fragmented regulatory landscape harms startups. The order tasks the Justice Department, Commerce Department, FTC and FCC with reviewing and potentially preempting state rules. Industry leaders and legal experts warn that the move could spark extensive litigation, extending uncertainty for smaller AI firms that lack resources to navigate conflicting state and federal demands. While supporters hope the order will spur Congress to craft a unified national framework, critics say it may delay clarity and burden innovators.Read more

Time magazine chose a collective of AI leaders—dubbed the ‘Architects of AI’—as its Person of the Year, highlighting the outsized influence of engineers, CEOs, and researchers from companies such as OpenAI, Nvidia, Google, Meta, and Anthropic. The selection underscores how artificial intelligence has moved from niche research to a daily fixture for millions, reshaping consumer habits, economic power, and geopolitical dynamics. Figures like Sam Altman and Jensen Huang are credited with building tools that now underpin everything from chat assistants to global chip strategies, marking AI as a central force in modern life.Read more

OpenAI touts its latest model, ChatGPT 5.2, as the smartest generally‑available AI, but early reactions on Reddit describe it as overly corporate, safe, and a step backward from previous versions. Users criticize the upgrade for feeling robotic, lacking spark, and prioritizing caution over creativity. While the model has just launched and many of its 800 million weekly users may still be testing it, the vocal dissent on the ChatGPT subreddit highlights a sharp divide between official claims and user experience.Read more

President Trump signed an executive order aimed at establishing a single federal framework for artificial intelligence and limiting the impact of state AI regulations. The order directs federal agencies to reduce the influence of state laws, creates an AI Litigation Task Force to challenge state statutes deemed inconsistent with federal goals, and targets Colorado’s consumer‑protection law on algorithmic discrimination. It also tasks the FTC, Commerce Department and FCC with issuing guidance and standards that could preempt state requirements, while carving out exceptions for certain lawful state regulations. Industry leaders and political observers note the order grants significant power to the White House AI and crypto czar, David Sacks, and reflects ongoing tension between federal and state approaches to AI governance.Read more

President Donald Trump signed an executive order titled “Ensuring a National Policy Framework for Artificial Intelligence.” The order creates a Justice Department task force to challenge state AI statutes that conflict with a forthcoming federal framework and directs the Commerce Department to withhold future broadband funding from states that enact what the administration calls “onerous” AI regulations. Supported by industry groups and White House advisers, the order also instructs advisers David Sacks and Michael Kratsios to draft legislative recommendations for a national AI policy, while carving out limited protections for child safety, data‑center infrastructure, and state procurement. State attorneys general and civil‑rights groups have criticized the move as an overreach that will likely face legal challenges.Read more

Google announced a new version of its Gemini Deep Research agent, built on the Gemini 3 Pro model and featuring an Interactions API that lets developers embed advanced research capabilities into their applications. The agent can handle large context prompts and is being integrated into Google Search, Google Finance, the Gemini app, and NotebookLM. Google also introduced a new benchmark, DeepSearchQA, to evaluate complex, multi‑step information‑seeking tasks. On the same day, OpenAI released GPT‑5.2, codenamed Garlic, positioning it as a direct competitor in the evolving AI‑agent landscape.Read more

Theo Baker, a senior at Stanford University, earned the George Polk Award after his reporting led to the resignation of Stanford president Marc Tessier‑Lavigne. His investigative work, which involved hundreds of interviews and extensive field work, has drawn the attention of major studios and publishers. Baker is now preparing to release a book that examines the money‑driven subculture of Silicon Valley venture capital and its impact on university students. The project, based on more than 250 interviews, promises to expose the ways young entrepreneurs are courted and exploited before they have fully formed ideas.Read more

Runway announced its new GWM-1 "world model" technology, asserting it can maintain coherent output for minutes at a time. The company framed the model as a step toward unifying diverse domains and action spaces under a single base system. While highlighting potential uses in film, television, advertising, robotics, physics and life‑science research, Runway also disclosed a partnership with CoreWeave that will employ Nvidia GB300 NVL72 racks for training and inference. The move positions Runway in a competitive AI arena where larger tech firms also pursue similar capabilities.Read more

OpenAI has launched its latest GPT-5.2 model, available through ChatGPT and the API. The new system emphasizes clearer self‑explanations, stronger multi‑step reasoning, and a vastly expanded memory that can track information across lengthy documents and multiple files. It also delivers higher reliability, with hallucinations reduced by about 30 percent compared with its predecessor. GPT-5.2 is offered in three tiers—Instant, Thinking, and Pro—and is being rolled out across OpenAI’s subscription plans, signaling a major upgrade for both casual users and professional workflows.Read more