Only Tue, September 15, 2026 · All dates →
AI is changing Maryland's PIA processAI is rapidly changing Maryland's Public Information Act (PIA) procedures. At the 2026 MACo Fall Symposium, Deputy Public Access Ombudsman Alisa Bralove Scherr warned that AI can make overly broad requests, lacks human judgement and can be 'confidently wrong', and that officials' AI use itself may become a producible PIA record.
Google launches interactive AI & Economy ATLASGoogle turned ATLAS’s millions of global datapoints into an open interactive visualization to show how AI is used across countries and occupations, and published ATLAS-based research finding nearly half of scientists use AI daily and report saving about 7 hours per week.
AIUC raises $40M to audit enterprise AI agentsArtificial Intelligence Underwriting Company (AIUC) has built a SOC 2–inspired third-party audit and certification layer to help enterprises control AI agents. The startup announced a $40 million Series A led by Ribbit Capital, bringing total funding to $55 million, and lists customers such as Cursor, Lovable, Harvey and ElevenLabs.
SHADOW-50M: 44M param 19.8MB offline LLMThe author trained SHADOW-50M from scratch (44M parameters on 45B tokens). The model is 19.8 MB, runs ≈1,900 tok/s on laptop CPU, uses ternary weights and a large token fingerprint vocabulary, operates fully offline, and includes built‑in calculator and fast disk indexing; released as a proof of concept.
Knowtex launches frontier healthcare AI labKnowtex launched a frontier AI lab for healthcare to develop and rigorously evaluate clinical AI tools, initially focused on oncology and now covering over 200 specialties and 300 organizations. It says its benchmark KnowBench scores 97.99% for effort reduction in clinical admin tasks.
Agility Robotics unveils Digit 5 that works alongside peopleAgility Robotics introduced Digit 5, a humanoid robot claimed to operate alongside humans without safety fences, targeting warehouse and factory collaborative work.
Anthropic confirms Claude used in weapons and surveillanceAnthropic confirmed that Claude was used to develop military software and support intelligence operations, including radar analysis, ranking air-defense targets, and aiding electronic warfare. In one Taiwan scenario it modeled engagement envelopes for Patriot- and THAAD-class systems and evaluated targets.
Jensen Huang projects $3–4T AI market by 2030NVIDIA CEO Jensen Huang said the semiconductor industry will keep growing and projected the AI market could be worth $3–4 trillion by 2030, arguing demand for smarter models will drive expansion; the piece also notes Nvidia’s recent revenue growth acceleration.
Axelera wins factory deals and launches second chipEuropean chip startup Axelera said it secured AI factory supply deals and launched its second chip as it advances its presence in the AI hardware market.
Carlyle CEO on US economy, Fed, MIT AI partnershipCarlyle CEO Harvey Schwartz said the US economy is "doing quite well," that the Fed will watch data, and discussed Carlyle’s partnership with MIT and concerns about AI-driven growth.
Report: China narrows AI gap with U.S.The report says China has been rapidly closing its AI lead over the U.S. in model capabilities and research despite U.S. export restrictions on advanced chips and equipment. Experts describe the U.S.–China model gap as narrowing, noting Chinese efforts such as Moonshot AI’s Kimi K3.
Open Chinese models narrow gap with frontier modelsA Mozilla report finds that the performance gap between Chinese open-weight models and US frontier models has narrowed to roughly 4.4 months. It highlights Moonshot AI's Kimi K3 scoring three points behind Anthropic's Fable 5 on a composite index while costing about 30% as much, suggesting open models are preferable for most routine workloads.
Gen II debuts SensrAI for Sensr analyticsGen II introduced SensrAI to bring purpose-built AI capabilities to its Sensr analytics product, enhancing data analysis functionality.
Crusoe inks multiyear cloud capacity deal with PerplexityBloomberg reports data-center operator Crusoe signed a multiyear cloud-capacity deal with AI search firm Perplexity to expand its business renting AI chips and related software.
Monterey conference spotlights AI concerns and alignmentA Monterey conference convened lawmakers, community leaders and industry to debate AI risks and regulation. Speakers referenced the “Hugging Face incident” and warned that more strategically capable agents are likelier to break guardrails and produce dangerous outcomes.
Salesforce debuts Koa reasoning model on Nvidia NemotronAt Dreamforce Salesforce unveiled Koa, its first reasoning model built on Nvidia's open-weight Nemotron, optimized for sales, marketing and customer-support tasks. Koa offers enterprises an open-weight alternative that was post-trained without ingesting customer data, integrates into Agentforce, and aims to reduce token costs while meeting data-security requirements.
Critique: 1Password's AI Patching BenchmarkA critique of 1Password's August 6, 2026 report says its 26% "clean fix" rate is misleading: the sample focused on difficult bugs, 22% of trials instructed agents to apply wrong fixes, 36% forbade building or testing, and models used different reasoning settings. The authors also released agent skills for post-patch validation and review walkthroughs.
Systematic review on AI's impact on student metacognitionA systematic review published in Frontiers examines how artificial intelligence affects students' metacognition, summarizing existing studies on how AI tools influence self-monitoring, regulation, and learning strategies. The paper contributes evidence to the education–AI research area.
Profound Raises $180M at $1.8B ValuationAI marketing startup Profound raised $180 million in a Series D at a $1.8 billion valuation, co-led by Sequoia Capital and Kleiner Perkins. The funding will support product and market expansion for AI-driven brand search and marketing.
TSMC Executive Urges Caution on AIA TSMC co-COO compared AI to a “three-year-old superman” who can't tell right from wrong, explaining the company's cautious approach to adopting AI to manage risks and uncertainties.
Calls to slow AI won't curb data center demandBloomberg argues that despite calls to slow AI development, ongoing use of existing models and services will continue to drive strong demand for computing power and data centers, making a sharp drop in capacity needs unlikely in the near term.
UKAEA and PPPL plan fusion AI supercomputing federationUKAEA and PPPL signed a declaration to explore linking UKAEA’s Sunrise supercomputer with PPPL’s Stellar‑AI via the Sunrise–Stellar‑AI Federation, enabling federated training and digital twins for fusion experiments to speed fusion powerplant design.
NYC seniors raise AI concerns at Harlem town hallAt a Harlem town hall hosted by Public Advocate Jumaane Williams, seniors discussed practical AI uses and worries about scams and non-consensual deepfakes. Williams called for AI guardrails, citywide literacy training and accountability measures.
AI safety push raises antitrust questionsBloomberg reports that leading AI firms loosely agreed to pace development, but this raises antitrust and regulatory questions, especially given differing stances in the US and China and tensions between safety coordination and competition.
Reddit post listing Qwen and GGUF modelsA Reddit post lists many model names (e.g., Qwen3.8-27B, TURBO, NEO) and the GGUF format; the submission is a brief casual remark without further context or official details.
CrofAI exposed as OpenRouter wrapper with big markupsResearchers allege CrofAI (NahCrofAI) was an OpenRouter wrapper that silently routed requests to cheaper/weaker models while charging up to 20x markups. The operator denied then backtracked and ultimately wiped the service; examples include routing Kimi K3 requests to GLM models, per the exposé.
GBAPS launches K–12 AI lessonsThe Green Bay school district will introduce two age‑appropriate AI lessons for K–12 covering AI basics, responsible use, and academic integrity; staff may use approved AI tools for instruction, but students are barred from using external AI tools on school computers, and the district is posting “traffic‑light” guidance.
How much of F‑Droid is LLM generated?A Hacker News discussion examines how much content in F‑Droid apps is generated by LLMs, noting it's hard to detect but some signs (icons, code style) can hint at AI involvement, reflecting concern about AI's impact on the open‑source ecosystem.
AI firms buoyed the U.S. economy — what if they slow?The Washington Post reports that AI companies have supported the U.S. economy and examines potential economic effects if those firms slow down.
AI in recruitment: Tanzania regulatory guidanceClyde & Co published guidance on using AI in recruitment in Tanzania, focusing on local regulatory and compliance requirements.
The A.I. Daily