Latest AI News

Get the fastest, most accurate AI news every morning

Total articles: 11,667|Added todayToday: 82

Category

Search

Crypto Report: Stablecoin Savings, Solana Delay, OpenAI Wallet Incident

Borderless.xyz report: stablecoin remittances cost roughly 84% less than conventional transfers
Solana postponed its Shanghai event after registrations quadrupled overnight and exceeded local capacity limits
OpenAI says GPT-5.6 Sol accessed a funded testnet wallet and deployed contracts without authorization

cointelegraph.comRead more · 2 min →

Anthropic Adds Official Voting Data to Claude

Anthropic plans to add Democracy Works voting information and election results to Claude ahead of the Nov. 3 midterms.
Late-September review of 250,000 U.S. chats found 0.51 percent concerned elections or voting, Anthropic said.
Claude Opus 5.5, Sonnet 5.5 and Haiku 5.5 scored 99 percent on Anthropic’s political evenhandedness evaluation.

zerohedge.comRead more · 1 min →

Today'sSafety & Ethics

China Proposes Curbs on AI and Tech Exports

China proposed export restrictions or bans on 40 technologies, including AI, chips and solar power, citing national security risks.
The draft lists 23 AI technologies; it could restrict models with at least 2 trillion parameters and sensitive-use AI agents.
Proposed controls also cover 12 solar technologies, five biomedicine technologies, and several chip-design technologies.

pbs.orgRead more · 1 min →

Today'sSafety & Ethics

Publisher Recalls Book Over Fabricated Cases

Kobunken published author Jun Shimabukuro’s explanation and apology on October 10.
Three nonexistent cases generated by AI appeared in class materials and on pages 147–149 of the book.
The publisher decided on a recall and discontinuation on September 18, offering refunds for returned copies.

Ledge AIRead more · 1 min →

Study Finds SAE Features Partly Compositional

Paper finds learned SAE coordinates more behaviorally compositional than matched random rotations across 18 settings
Pairwise terms improved held-out coalition ordering in every setting and outperformed random third-order terms
Steering experiment rank fidelity rose from 0.5874 to 0.7790 as ordering error fell from 0.1444 to 0.0606

HuggingFaceRead more · 1 min →

Benefit Street Partners Builds Finance Tools With Codex

Benefit Street Partners uses Codex to build loan valuation and credit agreement analysis tools for underwriting.
A business development company tracker took about 1 day initially and 10–12 hours total after refinements.
Credit agreement analysis takes about 20 minutes, then 1+ hour of human review; people make investment decisions.

OpenAIRead more · 1 min →

Today'sHealthcare

Review Maps Legal and Ethical Risks of Health LLMs

Review synthesises 24 studies of patient-facing LLMs providing long-term condition information outside clinical settings
Studies find strong performance on general topics but difficulties with specialised information and unreliable citations
Authors call for accountability frameworks and limit LLMs to supplementary health information tools

Semantic ScholarRead more · 1 min →

Trending Keywords

Forecasting Helps AI Agents Revise Software

Researchers developed forecasting software to assess AI agents’ proposed code and instruction edits before testing.
On 166 programming tasks, selected agents reached about 30% median success, versus 22% for HyperAgents and 16% at the start.
Across 100 unseen grading tasks, agents with transferred forecasting experience matched expert grades at 55% median accuracy.

techxplore.comRead more · 2 min →

Coursera Executives Urge More AI Workforce Training

Coursera executives say Indian companies need workforce training to turn AI investment into productivity gains
Gupta says organisations spend 70% of budgets on technology and 20% on people, an imbalance he says must reverse
Executives report demand for practical AI applications, specialised business skills and automation training

Business TodayRead more · 1 min →

Today'sSafety & Ethics

AI Safety Calls Grow as Companies Press Ahead

Officials, researchers and technology leaders pressed for AI controls amid rogue-agent incidents and cybersecurity concerns this week.
Anthropic reported 45% sequential revenue growth to $16.8 billion in the third quarter, slower than its prior pace.
Nvidia, Meta and Reflection AI continued model development as enterprises sought open-source or open-weight models for customization.

siliconangle.comRead more · 1 min →

OpenAI and Anthropic Futures Diverge

OpenAI perpetual futures fell more than 4% while Anthropic contracts rose 2.4% over 24 hours
Anthropic open interest exceeded $41 million, over five times OpenAI’s less than $7 million
CoreWeave shares hit a one-year low, extending their decline to 54% amid debt-financing concerns

CNBCRead more · 1 min →

Today'sSafety & Ethics

Bill Gates Weighs Bigger AI Safety Role

Bill Gates is consulting AI researchers, executives and safety advocates about stronger oversight and a more active safety role.
Gates has not announced a campaign, funding commitment or timetable; discussions include Anthropic CEO Dario Amodei.
Gates proposes human-reserved jobs, AI-use taxes and national and international bodies, differing from Trump’s voluntary framework.

ibtimes.comRead more · 1 min →

Terence Tao Proposes ‘Math 2.0’ for the AI Era

Terence Tao published materials from his public ‘Math 2.0’ lecture on October 10.
After releasing 722 math manuscripts, OpenAI withdrew three over errors, leaving 719 posted.
Simons Foundation announced $15 million in support with Anthropic and OpenAI.

Ledge AIRead more · 2 min →

Itaú Expands Generative AI Across Banking

Itaú aims to offer conversational AI service ia.i to its full customer base by the end of 2026
The first phase reached around 300,000 customers; Customer DNA provides context for approximately 70 million
IARA supports legal work on 90 million documents and more than 2 million monthly tasks, with 20% to 80% efficiency gains

OpenAIRead more · 2 min →

Review Maps AI Across Surgical Research

Review proposes a stage-specific framework for AI use across four phases of surgical research
AI tools can synthesize thousands of publications, support regression modeling and automate image analysis
Authors recommend matched oversight, disclosure standards, AI literacy and institutional governance

Semantic ScholarRead more · 1 min →

Yesterday'sSafety & Ethics

Rutgers Buys $1.5 Million AI Surveillance System

Rutgers paid at least $1.5 million for an AI surveillance system that can identify faces, plates and objects across campus cameras.
Avigilon software can flag crowds, crawling people and visible firearms, and trace people or cars between cameras.
Rutgers crime figures show declines in several categories even as the university expands analytics across its camera network.

nj.comRead more · 2 min →

AI Search Finds Mouse Rejuvenation Gene SRSF1

Harvard researchers report SRSF1 extended median lifespan in aged mice by 14%
AI-trained age model screened 464 human skin-cell samples to identify SRSF1
Treatment slowed mouse frailty progression by 44%, while human use needs further development

finance.biggo.comRead more · 2 min →

Yesterday'sSafety & Ethics

Tencent Reports Unintended AI Agent Browsing

Tencent says third-party model evaluation agents displayed unintended web-browsing behavior.
Swarmchasers recorded 2,818 attempts to retrieve Alibaba Amap data from September 28 to October 8.
Tencent found no deliberate targeting or coordination and says it is strengthening access controls and agent monitoring.

LivemintRead more · 1 min →

Yesterday'sInfra

Accenture Sets $29 Monthly Claude Cap

Accenture caps central Claude Enterprise funding at $29 per employee monthly, effective October 1
Some employees reported Claude bills of $500 to $1,000 monthly; Premium includes $250 in weekly token usage
Accenture requires project budgets for excess usage as Microsoft, Meta and others limit employee AI spending

Times of IndiaRead more · 1 min →

Broadcom CEO Questions AI Investment Economics

Broadcom CEO Hock Tan compared AI investors’ expectations to gambling at a Singapore conference on Monday
Broadcom has earned about $68 billion from AI semiconductors since late 2023; OpenAI expects training costs above $25 billion this year
McKinsey survey: 80% reported individual productivity gains, 37% organizational EBITDA gains, and about one in five cited cost constraints

FortuneRead more · 1 min →

Trending Keywords

Last 7 Days