Models, products and AI startups, near-duplicates collapsed to the most credible source
Anthropic says it blocked possible biological‑weapon effortsAnthropic says it blocked attempts that could have enabled biological‑weapon development, a matter reported by the press that highlights the company’s role in AI safety and misuse prevention.
Report: AI may worsen discrimination in US justicehumanrightsresearch.org published a report titled "Artificial Influence" examining how AI can contribute to discrimination in the U.S. criminal justice system, highlighting related policy and research concerns.
BofA: Ternus era emphasizes innovation, privacy and AIBank of America says John Ternus’s first major product launch as Apple CEO highlighted the company’s focus on product innovation, privacy, and personalized on‑device AI, noting the foldable iPhone Duo and on‑device/Private Cloud Compute capabilities.
Microsoft report on K–12 AI adoptionA Microsoft report outlines trends in K–12 AI adoption, summarizing how AI is being used across tools and curricula from early grades through high school.
Morris County to launch AI and Robotics academyMorris County Vocational School District plans to open an academy for artificial intelligence and robotics engineering to provide vocational education and boost regional skills in AI and robotics.
AI for infection prevention: surveillance and trainingThe article outlines how artificial intelligence can support infection preventionists (IPs) with surveillance, training, and patient safety, enabling enhanced monitoring and education to improve infection control effectiveness.
Ex‑DeepMind PR says lab banned public talk of extinction riskA former DeepMind spokesperson says the lab once prohibited public discussion of AI‑driven human extinction, while internally teams acknowledged that AI alignment remained unsolved.
Skild AI launches S1 robot foundation model learning from one videoSkild AI introduced S1, a robot foundation model that learns new, long‑horizon tasks from a single video without weight updates; the company reported a $100M annual revenue run rate, over 60 deployment partners, and collaboration with NVIDIA for training and deployment.
AI Boom Enters Power-Constrained PhaseBloomberg reports the AI boom is entering a phase where electricity may become the main constraint, pushing data centers toward stranded power; cutting compute per task and market-driven guardrails will shape the next wave of AI investment.
CISA issues advisory on China‑based attacks targeting AI firmsCISA released an advisory warning of China‑based attacks targeting artificial intelligence companies and urged the industry to strengthen defenses and adopt mitigation measures to protect AI assets and data.
Top AI stocks to research — Sept 10MarketBeat published a list of top artificial intelligence stocks to research, offering investors pointers on AI‑related market opportunities and names to consider.
AWS: Model‑agnostic PII detection with LLMsAmazon Web Services describes a model‑agnostic approach that uses large language models to detect personally identifiable information (PII) without relying on a specific model architecture, aiming to improve cloud data security and compliance.
NVIDIA: robotaxis are physical AI's breakthroughNVIDIA says the global robotaxi market could reach $400 billion by 2035 with over 6 million commercial vehicles. Scaling fleets requires massive compute; NVIDIA offers an open platform—from DGX training systems to simulation and in-vehicle computing—and says major commercial robotaxi programs use its stack.
3 ways Search can prep your raceGoogle outlines three AI-powered Search features to help runners: Canvas for personalized training plans, YouTube Music integration for playlists, and filtering tools to find suitable gear.
Using Codex and ChatGPT to find antimicrobialsCésar de la Fuente’s lab uses Codex and ChatGPT to search living and extinct genomes for antimicrobial candidates, demonstrating large models’ potential in biological discovery against drug-resistant infections.
Feyn open-sources promptable MultiMatte modelFeyn released MultiMatte, an open-source promptable image background removal model built on SAM 3 that uses alpha mattes to better handle hair and motion blur, reporting substantial metric gains on DIS5K and DUT-OMRON benchmarks.
Positron valuation soars in funding roundReuters reports AI chip startup Positron saw its valuation surge in a recent funding round, highlighting continued investor interest in AI hardware.
UMG and ElevenLabs launch AI music platformUniversal Music Group signed a multi-year licensing deal with ElevenLabs to build an AI platform letting users create remixes and new takes from licensed tracks; artists can opt in or out.
BNP: Excess AI Borrowing Could End Credit RallyBNP Paribas predicts the long corporate bond rally could break when tech companies flood markets with too much debt, suggesting AI-related borrowing growth may become a tipping point for credit markets.
Skild AI Hits $100M Revenue Run RateRobotics-software startup Skild AI reached a $100 million recurring revenue run rate 10 months after starting commercial use, indicating rapid growth and efficient commercialization.
Cognition Unveils SWE-2 Coding ModelCognition launched SWE-2, a coding model scoring 50.0 on FrontierCode 1.1 Main—within one point of Fable 5.1 while 64% cheaper; it was post-trained from the 2.8T-parameter Kimi K3 and uses single-run RL across reasoning levels to improve cost-performance.
Analysis: Claude Fable 5.1 writes more plainlyArena.ai analyzed tens of thousands of benchmark responses and found Claude Fable 5.1 writes more matter-of-factly but also more verbosely than Fable 5, indicating subtle stylistic shifts.
Dario and Daniela Amodei profiled in JPostThe Jerusalem Post profiles siblings Dario and Daniela Amodei as influential figures shaping the next frontier of AI, focusing on their roles and impact in the industry.
Data agent arrives in ChatGPT WorkChatGPT Work adds a Data agent that connects company data, uncovers insights, and builds interactive dashboards via natural language, making enterprise data more accessible to AI-driven analysis.
OpenAI’s Hugging Face hack triggers Senate probeA security incident involving Hugging Face and OpenAI has prompted a U.S. Senate probe, underscoring scrutiny of AI companies' security and compliance.
AI agents flood public services with requestsResearcher Chris Schmitz finds AI assistants have driven large increases in public-service submissions across 11 jurisdictions: 84 cases show spikes after ChatGPT, including UK housing complaints rising from 2,600 to ~7,000 and CFPB complaints growing 5x. The trend — dubbed “agentic flooding” — may persist and forces reconsideration of social-service design and policy.
OpenAI calls for national AI safety rulesOpenAI has shifted stance and publicly called for national AI safety rules, signaling a change in its approach to AI governance.
Meta, Anthropic and OpenAI shaken by pay and exitsReports say Meta, Anthropic and OpenAI are unsettled by billion‑dollar pay, departures, and fears over loss of control, reflecting talent and governance pressures at major AI firms.
Attorney uses AI ads and billboard for estate planningAn ABA Journal story describes an attorney using AI‑generated ads and a billboard to promote estate planning, leveraging tech for client outreach.
AI 2027 scenario reportThe AI 2027 scenario, authored by several contributors, predicts vast impacts from superhuman AI over the next decade, presents two endings—"slowdown" and "race"—and aims to use concrete, quantitative assumptions to spur discussion and alternative scenarios.
The A.I. Daily