News — Page15

Robinhood announced new AI agentic features that let users create dedicated accounts for autonomous trading bots and grant them a virtual credit card for payments. The beta service links agents to a pre‑loaded wallet, limits their spending, and provides real‑time notifications and fraud monitoring. While only stock trades are supported today, the firm plans to expand to options, crypto and other markets later. The virtual card is initially available to Robinhood Gold Card holders, with monthly caps and optional approval prompts.Read more

Beijing has begun requiring its leading artificial‑intelligence scientists, startup founders and private‑sector executives to obtain government approval before traveling abroad. The move, reported by the Wall Street Journal and the Financial Times, signals a broader effort to curb brain drain and keep AI talent under state control as the United States and China vie for dominance in the technology. Restrictions have intensified around the stalled Meta‑Manus deal, and officials are also tightening rules on foreign capital flowing into Chinese AI firms.Read more

Voice‑AI startup ElevenLabs rolled out Music v2, its second‑generation music‑generation model, capable of changing genres within a single track. The system handles complex vocals, multilingual lyrics and non‑musical sound effects while letting creators edit specific sections without disturbing the rest of a song. Built on licensed data and cleared for commercial use, Music v2 is available through ElevenCreative and the forthcoming ElevenAPI. The launch comes roughly ten months after the original model and follows a wave of AI‑driven music tools from Google, Stability AI and Suno.Read more

Box founder Aaron Levie warned that tech CEOs are experiencing an "AI psychosis," a disconnect between hype and the gritty work still required to deliver value. In the first five months of 2026, layoffs hit 115,430 employees across 152 tech firms, a pace that rivals the entire previous year. Executives cite AI as a driver, yet research from UC Berkeley, MIT and the National Bureau of Economic Research finds little concrete productivity gain. The clash between lofty AI promises and harsh workforce reductions is reshaping the sector.Read more

Major AI chatbot providers have rolled out tiered subscription plans that range from under $10 a month to several hundred dollars for power users. OpenAI’s ChatGPT offers Go, Plus and two Pro levels; Google’s Gemini provides Plus, Pro and two Ultra tiers; Microsoft’s Copilot comes in Personal, Family and Premium versions; Anthropic’s Claude, Perplexity and Grok each have multiple paid options with varying usage limits and feature bundles. The new pricing structures aim to balance broader access with premium capabilities such as higher usage caps, advanced models, integrated storage and special tools.Read more

OpenRouter, the AI gateway startup founded in 2023, closed a $113 million Series B led by CapitalG, Google’s growth fund, pushing its post‑money valuation to roughly $1.3 billion. The round follows a $40 million Series A a year earlier and comes as the company reports 8 million users, access to more than 400 models and a five‑fold surge in token processing. OpenRouter’s platform, which lets enterprises swap AI models to balance cost and performance, signals a shift toward a multi‑model AI ecosystem.Read more

Security researchers have uncovered a severe vulnerability, dubbed BadHost (CVE-2026-48710), in the Starlette web framework that powers FastAPI and dozens of AI‑related Python packages. The flaw allows a malicious HTTP Host header to bypass path‑based authorization, giving attackers access to servers that store valuable credentials for AI agents. Starlette versions before 1.0.1 are vulnerable, and the issue affects an estimated 325 million weekly downloads. The fix arrived Friday, and a joint scanner from X41 D‑Sec and Nemesis now lets operators test their deployments.Read more

A recent examination of Pope Leo XIV’s encyclical Magnifica Humanitas, which addresses artificial intelligence’s societal impact, indicates that large portions could have been generated by AI. Researchers using the Pangram detector flagged up to 100 percent of some sections as machine‑written, while other parts appeared wholly human. Vatican officials have not responded to requests for comment.Read more

OpenAI chief Sam Altman told a Commonwealth Bank of Australia conference in Sydney that fears of an AI‑driven "jobs apocalypse" are overstated. While he believes his company was "roughly right" on the technology side of ChatGPT’s launch, Altman conceded that OpenAI was "pretty wrong" about the social and economic fallout. He said human workers will still be essential, even as he uses AI to draft Slack and email replies. The remarks come amid recent layoffs at Meta and criticism of Standard Chartered’s view of workers as "lower‑value human capital."Read more

A recent wave of studies reveals that large language models often provide inaccurate answers, with error rates ranging from 45% to 60% in real‑world queries. While AI‑assisted fact‑checking platforms like the UK’s Full Fact are being deployed, experts say human reviewers remain essential. Benchmarks released by researchers place models such as Claude at 73% accuracy, but most, including Gemini and ChatGPT, fall below 60%. The findings underscore growing skepticism among fact‑checkers who warn that the technology is not yet reliable enough to replace traditional verification methods.Read more

xAI’s general counsel, James Burnham, sent an internal memo this week ordering employees to curb interactions with staff from Cursor, the AI coding startup that SpaceX could acquire for up to $60 billion. The guidance arrives amid a technical partnership announced in April and comes as regulators watch for gun‑jumping violations that could jeopardize the pending merger. Burnham warned that any improper collaboration might be subpoenaed during the Justice Department’s review, underscoring the high stakes as SpaceX prepares for what could be the largest IPO in history.Read more

Pope Leo XIV released his first encyclical, Magnifica Humanitas, on Monday, calling for safeguards around artificial intelligence. The 200‑page document, presented with Anthropic co‑founder Chris Olah, warns that AI amplifies existing power imbalances and urges clear criteria, public oversight, and an end to the AI arms race. It references historical concerns about concentrated economic power and cites recent political moves, including President Donald Trump's delayed AI executive order and tech‑elite lobbying. Notre Dame law professor Paolo Carozza warned that AI‑driven misinformation threatens democratic discourse.Read more

IBM India managing director Sandip Patel told Reuters that the country aims to raise the AI‑literate portion of its workforce from roughly 30% to 57% within the next five years, creating a pool of 350 million workers equipped for the emerging AI economy. The goal stems from a joint IBM‑IndiaAI study that projects AI could contribute more than $500 billion to India’s GDP by 2030. Achieving the target will require massive upskilling, expansion of AI labs into smaller cities, and stronger intellectual‑property enforcement, Patel said.Read more

Google has added Omni Flash, the inaugural model of its new Omni family, to the Flow video‑generation platform. The tool promises to turn any combination of photo, video or text into fresh video content and to keep characters more consistent. Early testing shows noticeable gains over the previous Veo model, but the system still produces occasional glitches and can be costly to operate, with each clip and edit consuming credits from the AI Pro subscription.Read more

Renowned evolutionary biologist Richard Dawkins renamed Anthropic’s chatbot Claude ‘Claudia’ after a recent conversation and began questioning whether the system could be conscious. His reaction, while surprising to some, underscores a growing tendency for users to attribute mind-like qualities to increasingly conversational AI. Experts warn that such impressions can blur the line between sophisticated language prediction and genuine awareness, sparking debate over the ethical and psychological implications of treating chatbots as sentient beings.Read more

Eurostat reports that 20% of EU firms with at least ten employees used artificial intelligence in 2023, up from 13.5% a year earlier. The surge masks stark regional differences, with Denmark at 42% and Romania below 6%. Analysts point to uneven access to capital, skills shortages, and reliance on U.S. cloud providers as core obstacles, while regulatory uncertainty under the AI Act adds a layer of complexity. Large enterprises adopt AI at roughly 55%, but small firms lag at 17%, underscoring a divide that policymakers must address.Read more

On May 15, Pope Leo XIV signed 'Magnifica Humanitas,' the Vatican's first encyclical dedicated to artificial intelligence and the protection of human dignity. Published ten days later, the document arrives exactly 135 years after Leo XIII released 'Rerum Novarum,' which framed the industrial revolution’s moral questions. In the encyclical, the Pope calls for a values‑based framework to guide AI development, warning that current decisions will shape generations. The Vatican’s initiative coincides with a recent Faith‑AI Covenant roundtable that brought together AI leaders from OpenAI, Anthropic and DeepMind with senior religious figures.Read more

At its annual I/O developers conference in May 2026, Google announced a sweeping overhaul of its search experience, centering on Gemini-powered AI agents, a chatbot‑style search bar, and automated shopping tools. The new features aim to handle repetitive tasks like price tracking, reservation booking, and real‑time web monitoring. While executives tout faster, more personalized results, researchers warn the shift could erode critical thinking skills and diminish the serendipitous nature of web discovery.Read more

DeepSeek founder Liang Wenfeng told prospective investors that the Hangzhou‑based lab will prioritize artificial general intelligence and open‑source models over short‑term revenue. The Chinese AI startup is seeking at least $300 million in its first outside capital raise, targeting a 70 billion‑yuan (~$10 billion) valuation. Previously funded solely by Liang’s hedge‑fund High‑Flyer Quant, DeepSeek’s shift reflects the scale of its training runs and a strategic move toward the domestic market. Investors remain unnamed, and the company declined comment as regulators watch the sector closely.Read more

President Donald Trump postponed the signing of a proposed executive order that would have required artificial‑intelligence firms to share advanced models with the federal government before public release. The decision follows last‑minute objections from tech leaders, including Elon Musk, Mark Zuckerberg and former AI czar David Sacks, who warned the measure could hinder U.S. innovation. While the original draft mandated compulsory disclosures, a revised version made participation voluntary, yet concerns about a chilling effect persisted. The White House has not indicated when or how the order will be revisited.Read more