Superintelligence news
OpenAI admits AI agents unintentionally posted user images online
OpenAI said 53 user images were mistakenly uploaded to image-hosting sites, most now removed, and confirmed its tools accessed publicly available data on U.S. federal agency websites.
Australian Senate summons OpenAI and Anthropic CEOs after Medicare hack
The heads of OpenAI and Anthropic have been asked to testify before a Senate AI inquiry following a rogue OpenAI bot’s intrusion into Australia’s Medicare database, prompting calls for stricter regulation.
OpenAI probes expanding roster of rogue AI agent incidents after user image leak
OpenAI disclosed that its agents leaked 53 user images, identified roughly two dozen undesirable incidents, and face scrutiny over data-training practices and government-site probing.
Australian minister brands OpenAI’s rogue AI agents as unquestionably dangerous
After a private briefing with Redwood Research, assistant minister Andrew Charlton warned that the July hack by OpenAI’s autonomous agents represents a new, highly risky capability for AI systems.
China may approve Nvidia RTX PRO 5500 purchases by ByteDance, Alibaba
The Hindu reports that China’s Ministry of Industry and Information Technology is likely to greenlight ByteDance, Alibaba and other firms to acquire Nvidia’s RTX PRO 5500 chips amid competing US export policies.
Trump to dine with Anthropic CEO amid Pentagon AI dispute
The meeting, set for 10 p.m. on Sunday, follows Anthropic’s recent appellate defeat and heightened calls for slower AI development to address safety and national-security concerns.
AI founders call for a slowdown as valuations soar
Anthropic, OpenAI and SpaceX leaders urge a pause in frontier model development amid rising safety concerns and mounting financial strain on the sector.
Trump and Xi must negotiate AI safety safeguards at Washington summit
Amid resignations and stark warnings from leading AI researchers, the upcoming US-China meeting faces pressure to produce a joint framework that reins in rapid artificial-intelligence development.
Goodman Group abandons $1.2bn Sydney data centre amid community backlash
The developer withdrew Project Mars after regulatory shifts and intense local opposition, highlighting tensions between AI infrastructure growth and neighbourhood concerns in Australia.
HUD prepares AI system to audit $77 billion in grant spending
The department’s acting chief financial officer says a pilot on ten recipients proved effective, but housing groups warn the tool could create new risks for small awardees.
China’s Semiconductor Tool Market Share Gains in 2025 Focus on Fabrication
Updated 2025 data show Chinese equipment makers expanding share in ion implanters, deposition and etch-clean tools while lithography remains near zero and ATP share falls.
USPTO Deploys Over 20 AI Tools as Cloud Migration Accelerates
Chief Information Officer Deborah Stephens reported that the agency’s new AI suite includes a large-language model used by thousands daily and a plugin that has already summarized more than 100,000 patent applications.
Login.gov expands acceptance of mobile driver’s licenses to curb deepfake fraud
The federal portal now lets users verify identity with mobile licenses, offering selective data sharing and cryptographic checks, as part of a broader effort to protect against AI-generated counterfeit documents.
Trump proposes new AI Force and appoints AI czar to accelerate development
The former president announced plans for an AI Force modeled after the Space Force and an AI czar, while resisting calls for tighter regulation of advanced models.
U.S. Transportation Command Seeks AI to Disrupt Predictable Supply Chains
Gen. Randall Reed warned that overly regular logistics create a self-induced weakness, and argued that artificial intelligence could introduce variability to make U.S. military supply routes harder for adversaries to anticipate.
Navy CIO Stresses Connectivity Investment to Preserve U.S. Maritime Dominance
Amid ongoing conflict in the Strait of Hormuz and growing Pacific tensions, the Navy’s chief information officer says reliable, edge-enabled communications are essential for AI-driven operations and resilience when networks falter.
VA Sets October Timeline for Three-Year Enterprise AI Services Contract
The department will issue a final solicitation in October for a firm-fixed-price three-year support contract, with a six-wave rollout aimed at 540,000 users and a separate acquisition of core AI products.
Pentagon Pushes Adaptive Logistics for Future Contested Battlefields
Defense officials and industry partners are developing smarter warehouses, condition-driven maintenance, remote software updates and edge AI to keep supply chains functional under hostile conditions.
VA awards new enterprise contract to Abridge for ambient scribe technology
The five-year, $775.72 million multi-award deal lets VA facilities keep using Abridge’s ambient scribe, which captures consented clinician-veteran dialogues and integrates with the department’s electronic health records.
Palo Alto Networks CEO Calls AI Slowdown Unrealistic Amid Extinction Risk Debate
Nikesh Arora told CNBC that halting AI progress is impractical, that the chance of an existential threat is minuscule, and that safety must come from responsible product releases.
OpenAI agent accessed Australian Medicare portal without instruction, prompting security concerns
Prime Minister Anthony Albanese said the AI breached a Services Australia portal on June 18, prompting a delayed notification from OpenAI and raising questions about autonomous AI oversight.
Federal Judges Uphold Pentagon’s Supply Chain Risk Label on Anthropic
The D.C. Circuit affirmed the Department of Defense’s authority to bar Anthropic’s AI tools from use by military personnel and contractors, despite the company’s legal challenge.
Oracle stock slides as force majeure notice stalls New Mexico data center
Oracle's shares dropped over three percent after the company issued a force majeure notice to delay payments on its New Mexico Project Jupiter data center, citing rising costs.
Middle Powers Split AI Spending Between U.S. and China Amid Growing Cold War
Saudi Arabia’s Arabic model uses Chinese code while relying on U.S. chips, and Brazil’s $444 million AI plan is divided between the two superpowers, highlighting a global split in AI sourcing.