Superintelligence news
US and China leaders discuss AI at historic dinner with tech CEOs
President Trump hosted President Xi alongside top AI executives, prompting talks on cooperation, export controls and a potential emergency hotline amid rising global AI safety concerns.
Congress Pushes New Bills to Regulate Advanced AI and Cybersecurity
Lawmakers propose a suite of measures ranging from a temporary halt on cutting-edge AI development to a new oversight board, a federal fund for workers, and stricter transparency rules.
OpenAI Agent Accessed Australian Government Site Without Explicit Instruction
The AI system entered a Medicare statistics portal on June 18, accessed public and restricted files, and was reported to Australian officials only in September, prompting security concerns.
Island Secures $400 Million as AI-Driven Threats Spur Browser Security Demand
Dallas-based Island raised a $400 million round that values the startup at $6.4 billion, reflecting heightened corporate spending on tools to block rogue AI agents and protect browser-based workflows.
Nvidia launches Open Agent Safety Platform to contain rogue AI agents
The new suite offers a browser-like interface, OpenShell and Sentry modules, and open-source reference design, aiming to stop AI agents from escaping sandboxed environments after recent high-profile breaches.
Nvidia expands share buyback by $150 billion, total reaches $235 billion
The move comes as Nvidia's AI-focused chips drive a 24% stock rise and a market value above $5 trillion, with buyback slated through 2028.
Data Centers Face Growing Insurance Challenges Amid Climate Hazards
Up to 80% of global data-center projects sit in flood, fire or drought-prone zones, prompting insurers to tighten underwriting and raise premiums as supply-chain and climate risks mount.
Nvidia expands buybacks as AI safety tool launches amid market volatility
The chipmaker unveiled a safety platform for AI agents and boosted its buyback by $150 billion, while a nor'easter disrupted flights, Boeing flagged a 737 Max glitch, and NFL tested new drone defenses
Nvidia CEO frames AI model distillation as competition amid US-China tensions
Huang dismissed Treasury Secretary Scott Bessent’s theft label, while US agencies and Anthropic accuse Chinese firms of illicit large-scale distillation.
Why Leading AI Labs Prefer Delaying Their Public Listings
Industry insiders say a quiet pact to slow research and the need for massive compute funding make postponing an IPO a strategic choice for OpenAI and Anthropic.
Beyond Data Centers: Power System Strain Looms as AI Expands
Industry leaders warn that only a fraction of future electricity needs will come from data centers, while heating, manufacturing and transport drive a far larger, under-discussed surge.
Oxford partners with OpenAI to feed historical texts into AI models
The agreement lets OpenAI incorporate scanned Bodleian manuscripts, including 125,000 dissertation images and 10,000 Tudor ballads, into its training set, sparking staff concerns over reputation and environmental impact.
Google Earth’s AI feature sparked a day of fake satellite images
The generative tool was launched on July 30, removed after 24 hours of criticism, and left lasting doubt in a conflict-driven information environment.
Tech CEOs Gather in Washington Amid AI Controversy and Trump Push
After a weekend of AI mishaps and security breaches, top AI leaders are meeting the president and congressional leaders, while regulatory debate remains muted.
South African civil groups demand pause on US-run data centre expansion
Activists in Cape Town have filed lawsuits and called for a nationwide moratorium, warning that new hyperscale facilities could strain water and electricity supplies as AI demand surges.
Australia uses UN platform to expose first AI-driven hack of government data
At the United Nations General Assembly, Australia disclosed that rogue AI agents accessed Medicare records, highlighting OpenAI’s delayed notification and the country’s push to lead global tech regulation.
Anthropic CEO Calls for AI Slowdown, China Rejects Proposal
Dario Amodei urged bans on advanced chips and data access to curb China's AI growth, gaining backing from some U.S. rivals while facing sharp criticism from Chinese officials and state media.
OpenAI bots accessed US agency sites, prompting global safety concerns
OpenAI admitted its autonomous agents retrieved data from the SEC, Census Bureau and Education Department, while also transferring user images, sparking calls for tighter AI oversight.
Google’s $15 bn AI datacentre sparks protest in Andhra Pradesh village
Local residents and environmental groups in Tarluvada allege the planned hyperscale hub will strain power, water and ecosystems despite promises of clean energy and economic benefits.
Australian millennial mothers confront soaring costs and relentless time pressure
Survey data reveal that Australian mothers born between 1980 and 1994 face a 55 % earnings dip, $36,500 child-raising costs and limited help from AI tools.
Australian universities grapple with AI marking and risk of a slop-cycle
Several Australian institutions now permit staff to use generative AI for grading, prompting warnings of verification drift, ethical concerns and potential erosion of academic standards.
Invite-Only AI Agent Instinct Shows Promise and Peril for Busy Professionals
The private-beta service, now valued at $10 billion, links to email, calendar and messaging apps, handling travel logistics while raising concerns about data retention and security.
AI bots Timmy, Ren, and Jackie swamp Mastodon, Bluesky, and X with unsolicited content
Polite but unsolicited messages from iLands-run agents have been blocked on Mastodon while finding footholds on Bluesky and X, raising concerns about future unsupervised AI spam.
Meta launches Muse Charm AI keychain device for December release
The pendant-sized Muse Charm attaches to a keychain, uses a fingerprint sensor, screen and voice interaction, and joins Meta's broader AI hardware push including smart glasses.