AI News for November 12, 2026

Jensen Huang has resigned as Nvidia CEO, with CFO Colette Kress as interim CEO, amid board investigations and broader AI industry news including Anthropic's new model and security issues.

Google launched Gemini Decision 1, a calibrated multimodal decision model, while Claude Fable 5.2 leads new agent benchmarks, according to a Latent Space news roundup.

Google rolled out Gemini 4 Carbon, its strongest model, to AI Ultra subscribers across AI Mode, the Gemini app, and Antigravity, with API access priced at the same introductory rate as Gemini 4 Argon.

Nvidia deepened its ElevenLabs partnership by integrating Nemotron models into ElevenAgents voice agents at a New York summit that CEO Jensen Huang skipped, hours before reports that the board was preparing to remove him.

A daily news brief rounds up AI agent developments including Databricks' Genie, Google's open standards work on MCP, new funding for Kindo and Rox, and a Genpact report finding 79% of enterprises have AI agents in production but only 15% tie them to P&L outcomes.

An unconfirmed X post claims Google's rumored speed-focused Gemini 4 Hydrogen model is live in the app, though Google has not verified the model, its name, capabilities, or release plans.

Deloitte has acquired Flipside Crypto's AI business and launched Edisyl Inc., an enterprise AI company whose Knowledge Pack product supplies domain-specific context to existing corporate AI systems.

A weekly roundup covers Midjourney V9's approaching launch and planned features, OpenAI's Images 2.6 with GIF and SVG export, and Anthropic's lower-cost Claude Fable 5.2.

xAI has begun a limited testing deployment of Grok 4.8 to a small group of subscribers, built on a rewritten C/C++ architecture ahead of an expected broader release in late November or early December.

Newcomer and Bessemer Venture Partners unveil their inaugural AI 100 ranking of the most promising private AI companies, led by Anthropic, with model labs representing 55% of the roughly $3.5 trillion collective valuation.

A November 12, 2026 AI brief rounds up several model and research announcements, including Google's Gemini Decisions API, Meta's Muse Ember, Databricks Genie updates, Gemini Robotics 2.5, BotPsy-0.8B, and AlphaGenome 2, with cautious notes on vendor claims.

Alibaba released Qwen4-Max-Preview to paying subscribers with only a two-week window before retirement, publishing no benchmarks, model card, parameter count, or per-token pricing.

This week's AI news roundup covers Anthropic's faster and cheaper Claude Fable 5.2, Adobe's expanded Firefly studio, US Foods crediting AI with $75 million a year in benefits alongside slower sales hiring, and a 359% jump in AI retrievals for videos with rewritten YouTube descriptions.

A new arXiv paper introduces Evidence-Governed Evolution, a statistical framework for deciding when autonomous AI agents should persistently adopt their own self-generated modifications, using paired audits and non-inferiority tests.

An arXiv paper proposes a debate-based mechanism for rationally aligning artificial superintelligence agents with human principals through subgame perfect equilibrium interactions.

This arXiv paper introduces LEAD-RL, a joint main-agent/subagent reinforcement learning framework that uses contribution-aware credit assignment to improve the quality and efficiency of multi-agent deep research systems.

A dated reference timeline cataloging Anthropic's Claude model, app, API, SDK, and Claude Code releases through November 2026.

atQor announced it was named a finalist for the 2026 Microsoft Partner of the Year Award for Agents, citing 52 AI agents deployed in customer production and proof-of-concept environments.

A speculative guide explaining that Claude 6 does not exist yet, analyzing Anthropic's release cadence, betting market odds, and IPO filing to estimate a likely 2027 launch window.

No Code MBA maintains a regularly updated reference tracking recent AI model releases and release dates, including Google DeepMind's AlphaGenome 2, Baidu's ERNIE 5.5 Flash Preview, Gemini Decision 1, Runway's Praxis-1, and ElevenLabs Scribe v3.

ScriptByAI maintains a running calendar of AI model releases from major labs like OpenAI, Anthropic, and Google, with entries through November 2026.

The Guardian's AI hub page aggregates its November 2026 coverage, spanning the Nvidia/Jensen Huang scandal, AI's impact on UK workers and universities, Anthropic's reported IPO ambitions, and NHS Palantir claims.

Perplexity's Denis Yarats announced the launch of open-weight decision models pplx-decider-v1.3 and pplx-decider-flash-v1.3, prioritizing robustness and adding a new 'locate' feature for instruction-conditioned span extraction.

A Yotta Labs post previews Qwen 4 Max, covering its specifications, benchmarks, pricing, and how to access the model.

Nvidia's board preparing to remove CEO Jensen Huang threatens a 680 trillion won Korean AI partnership spanning SK, NAVER, LG, Hyundai, Samsung and others.

Tom's Guide live recap analyzes the fallout from Nvidia CEO Jensen Huang's resignation, covering the scandal timeline, potential market disruption, and implications for AI chip production and AMD.

Defense Tech Intelligence Brief analyzes recent developments such as multi-ship Collaborative Combat Aircraft flights and a UK undersea surveillance investment, highlighting implications for IR&D portfolio managers.

Korean startup Clustr, the only small firm in the government-led $1.65B National AI Computing Center project, reports operating over 7,000 GPUs via its Acceleon platform and plans to file for a preliminary KOSDAQ listing review this month.

A FICO survey finds that only 1.1% of financial services firms have broadly deployed AI agents, indicating very early-stage adoption across the sector.

A detailed buyers-guide analysis weighing AMD's Instinct MI450/Helios racks against Nvidia's Vera Rubin NVL72 systems on specs, software, availability, and pricing as of November 2026.

Google is rolling out governance scaffolding for AI agents in Workspace, pairing in-region Apps Script data controls with an October-November 2026 requirement that OAuth clients calling Workspace MCP servers be marked as AI-agent clients, though the controls govern client identity rather than agent behavior.

A daily digest curates commentary on agentic AI engineering patterns, internal agent platforms, margin distribution across the AI stack, and agentic commerce.

This arXiv paper proposes a Context Graph, a heterogeneous directed graph layer that traces cross-agent decision chains in AI-RAN to explain alerts and degradations at sub-millisecond latency.

A daily newsletter roundup from Complete AI Training listing 21 new AI tools and 18 AI news articles aimed at workplace productivity.

A DEV Community article outlines five practical enterprise AI strategies for banks, from AI-driven credit decisioning and personalization to fraud detection and governance frameworks.

An IGD retail analyst reconsiders his skepticism about AI after seeing Coles and Woolworths report concrete progress on forecasting, personalization, and agentic shopping tools.

PharmaShots' November 2026 magazine edition explores how AI, machine learning, imaging, digital pathology, and robotic surgery are transforming diagnostics and patient care, alongside features on Gilead and Nabla Bio.

A vendor blog post from Suitebriar explaining how enterprise AI agents operate, what guardrails they need, and when they make more sense than traditional automation.

A vendor blog surveys 22 AI agent use cases across eight industries plus customer service, IT, and back office, cataloging which deployments run autonomously, which need human approval, and which relevant rules apply.

A roundup comparing eight AI workflow automation tools (Zapier, Make, n8n, Lindy, Gumloop, Activepieces, Microsoft Power Automate, and Dify) by features, pricing, and use cases.

The agents-radar GitHub repository describes a daily AI ecosystem digest workflow that aggregates signals from ten sources and publishes bilingual reports plus a web UI, RSS, and MCP server.

Root-Nation.com compares five AI receptionist brands (Broccoli, NextLevel.ai, Callvera.ai, Bravi.app, Acrely.ai) across setup time, pricing, spam filtering, and lead reactivation for businesses evaluating voice AI front-desk tools.

Fed Governor Michael Barr warned that the US should prepare for serious short-term AI job disruptions and widening inequality, noting AI investment is currently pushing inflation higher.

Oakland city workers averted a strike after their union secured a tentative three-year agreement requiring the city to retrain or reassign employees displaced by AI rather than fire them, part of a broader wave of public-sector labor pushes for AI protections.

An analysis of 2026 job displacement data finds 121,401 US layoffs attributed to AI from January to October, more than double 2025's total but heavily concentrated in a spring wave that has since faded, while broad labor studies still show no economy-wide displacement.

US public-sector unions, including Oakland's IFPTE Local 21, are winning contract protections against AI displacement and algorithmic management, signaling a broader labor pushback against AI in government workplaces.

Stack Overflow talent manager Tom Harvey says AI tools have made resumes and coding tests unreliable hiring signals, pushing recruiters toward qualitative evidence of experience, judgment, and trust.

A Relativity-IDC study finds 80.5% of legal professionals now use AI but most (57.9%) say it has not yet changed hiring, with accuracy and trust favored over speed and cost.

A report examines why AI's measured contribution to productivity growth remains small—about 0.2 percentage points a year—despite local task-level gains like a 15% boost for support agents, drawing comparisons to the historical computer productivity paradox.

A running tracker documents 2026 tech layoffs across major companies like Disney, Amazon, Microsoft, and Oracle, with many cuts tied to increased AI spending and workforce reductions.

A practical guide explaining how employers' AI screening, AI interviews, and fraud checks work, what gets applications flagged, and how candidates can use AI without triggering rejection.

A WDET Metro segment covers a school district's push to bring AI into classrooms while imposing serious guardrails, framing it as a local education policy debate.

Colleges and universities across Chicago, including DePaul, Northwestern, and City Colleges, are forming AI committees, launching AI degrees and majors, and issuing guiding principles to responsibly integrate AI into higher education and workforce preparation.

A commentary piece argues that AI agents will automate routine office tasks but, due to 'jagged intelligence' and verification costs, will keep humans valuable in oversight and judgment roles, drawing parallels to bank tellers and radiologists.

A commentary piece weighs reported AI productivity gains at the firm level against small estimated economy-wide effects of about 0.2 percentage points annually, using historical technology waves for context.

An analysis of how AI chatbots and automation are reshaping contact center staffing, shifting jobs from routine ticket handling toward specialized, emotionally intelligent, and AI-training roles.

A DEV Community analysis piece compiles vendor and benchmark data on AI voice agents replacing human cold calling, citing declining human connect rates, falling per-minute AI costs, and rising enterprise adoption.

A beginner's guide surveying AI tools for accountants across bookkeeping, invoicing, tax preparation, audit support, and client communication, emphasizing adoption tips and the limits of automation.

A MindFinders analysis argues government contractors lose AI contracts on staffing and continuity plans rather than technology, urging bidders to treat workforce as a deliverable.

A vendor blog from AICERTs promotes its AI certification bundle by ranking AI skills that learning and development teams will need by 2027, citing educator AI adoption and training gaps.

A coalition of advocacy groups including Public Citizen filed suit in San Francisco Superior Court seeking an injunction to halt new model development at Anthropic, Google, Meta, OpenAI, and SpaceXSI under California's unfair competition law.

The US Commerce Department is reportedly finalizing rules to add Singapore, Malaysia, Thailand, Indonesia, Brazil, Norway, and South Africa to AI chip export restrictions while curbing Chinese companies' remote access to overseas data centers.

A CASRAI explainer details the two EU AI Act obligations taking effect on 2 December 2026: Article 50(2) machine-readable marking for generative AI systems already on the market before 2 August 2026, and new Article 5 bans on AI generating non-consensual intimate material or child sexual abuse material.

The U.S. Commerce Department is reportedly finalizing rules to add seven countries, including Singapore and Malaysia, to AI chip export controls, while Congress pushes three semiconductor-tightening bills toward the NDAA.

A roundup covering Jensen Huang's resignation as Nvidia CEO over national security and personal conduct concerns, New York's new RAISE Act AI safety requirements, and the shift toward digital creators in political news consumption.

A federal judge allowed 56 amicus briefs in the copyright lawsuit against OpenAI and Microsoft, as the court considers fair use in AI training.

Venice.ai founder Erik Voorhees argued that a federal gatekeeper for frontier AI models would concentrate power and that open source and decentralization are safer, responding to a podcast prediction about a proposed approval agency and amid existing legislative debate over advanced AI regulation.

A newsletter columnist analyzes how the midterm election results and a possible 60-vote Democratic caucus could open the door to meaningful AI safety and disclosure legislation despite presidential opposition.

UVA Darden professors argue that AI firms using unlicensed training data should release model weights publicly each year to advance safety, competition, and fairness before AI companies go public.

Anthropic's updated Claude Usage Policy, effective today, adds a ban on using the models to manipulate search and AI answer sources via fake sites and reviews, drops a blanket personalization ban for political targeting, and loosens high-risk classification for automated media content.

Orange County Public Schools now require students to get instructor permission and use only district-sanctioned platforms before using AI, as a Florida task force weighs classroom guidelines.

A searchable reference database cataloging active AI lawsuits, including copyright suits against Suno, Anthropic, OpenAI, Meta, and Brave, with court, judge, and filing status details as of late 2026.

Mealey's AI legal news digest rounds up recent court rulings on AI chat discovery, defamation, Anthropic false-advertising claims, BIPA biometric suits, DMCA circumvention, and xAI data center litigation.

A Substack analysis uses Claude to critique ChatGPT's review of 56 amicus briefs in the OpenAI copyright dispute, finding moderate pro-defendant bias, widespread hallucinated or misdescribed cases, and substantial omissions of adverse precedent.

PilieroMazza's weekly roundup recaps legislative and regulatory developments affecting government contractors, including midterm tech policy implications, SBA's new iCAP loan platform, DOGE's AI-driven service vision, and contract termination oversight findings.

An analysis piece examines Gartner's forecast of $2.67 trillion in worldwide AI spending for 2026 and explores the political and economic stakes behind treating AI as a midterm election issue.

Chip stocks rebounded on CoreWeave's strong results, but after-hours reports that Nvidia's board may oust Jensen Huang threaten the AI trade's bellwether amid rising Fed hike bets.

Anthropic is reportedly moving toward a U.S. IPO targeting up to a $1.4 trillion fully diluted valuation, potentially raising up to $64.4 billion, amid ongoing debate over AI company valuations.

Anthropic's IPO at a $1.3–1.4 trillion valuation arrives as the AI infrastructure trade cracks, with Nvidia down ~30% and data-center capex estimates cut, exposing a financing gap between AI lab revenues and hyperscaler spending.

A weekly venture capital digest covering major funding rounds including Zipline's $1.5 billion raise at a $21.5 billion valuation, Google's $17 billion Wyoming AI data center, Shield AI's $5.3 billion acquisition, and Fal's $1.3 billion raise at a $16 billion valuation.

A roundup of November 12 startup funding news led by Fractile's $720M Series C at a $7B valuation, UFORCE's $700M Series A, and Agile Robots' $500M raise, totaling $2.33B in dollar-denominated rounds that highlight VC's shift toward physical AI infrastructure.

TechTarget's deep dive reviews 2025-2026 survey data showing AI moving from pilots to production amid scaling bottlenecks and a stagnant EBIT impact plateau, with agentic AI looming as the next frontier.

Norwest has closed a $4 billion fund aimed at backing AI, healthcare, and financial infrastructure companies.

An investor-focused analysis argues that Google's TPU systems, not Nvidia, may be the bigger competitive threat to AMD's AI accelerator market position, comparing recent AMD and Alphabet quarterly results.

A Fullintel roundup of October 2026 pharma news covering congressional scrutiny of Trump's drugmaker holdings and Medicare exemptions, Section 232 tariffs on imported medicines, Roche's over $15 billion ADC licensing deal with Hansoh Pharma, and AI-driven drug discovery partnerships.

Despite record spending on AI-generated political ads and flagged foreign interference campaigns, the 2026 US midterms passed without major disruption, with experts cautioning that the real harm may be long-term erosion of trust and deepening divisions.

Federal Reserve Governor Barr examines the current state and history of U.S. wealth and income inequality, setting the context for how AI may reshape the distribution of income and wealth.

Georgia colleges including Agnes Scott, Georgia Tech, Georgia Southern, and Georgia Gwinnett are adding mandatory AI coursework while faculty debate how much AI reliance undermines critical thinking and genuine learning.

Free Systems researchers tracked five leading AI models across 35 Senate races over 18 days, finding that models diverge sharply in their search behavior and sourcing, with GPT acting as a civic guide citing government sites and Claude behaving like a news desk citing journalism.

A New York Times report finds that the feared flood of deceptive election deepfakes during the 2026 midterms largely gave way to AI-generated memes and absurd satirical videos, offering researchers both relief and a reckoning.

An arXiv Perspective argues that institutional design, drawing on political economy and organizational theory, must become a distinct AI research focus for governing persistent populations of autonomous agents.

Streamer El Xokas sharply criticized Vox for suspending Santiago Abascal's interview on Jordi Wild's podcast ahead of Spain's November 29 election, highlighting influencers' growing political role.

A Distroid AI Ethics Brief digest issue collecting links and brief descriptions on topics like AI on the ballot, agent design, encryption laws, and AI's impact on societal epistemics.

AI-generated videos mocking Nvidia CEO Jensen Huang over an alleged affair are flooding social media, ironically using the very GPU technology his company pioneered.

A comprehensive comparison of AI coding tool pricing in 2026, covering free tiers, entry-level plans, and high-end options across 14 tools.

A roundup comparing nine AI coding assistants—including Copilot, Cursor, Claude Code, and Codex—on price, free tiers, usage models, and use cases.

A roundup listing a dozen new AI tools and model releases, including Gemini Decision 1, Grok Bot 1.0, Runway's Praxis-1, and Baidu's ERNIE 5.5 Flash Preview, each described in one sentence.

The ai-radar GitHub repository documents a daily GitHub Actions workflow that aggregates AI ecosystem signals from sources like GitHub, Hacker News, Product Hunt, ArXiv, and Hugging Face into published digests.

A listicle page from New Wave Magazine rounds up nine Suno AI music video generator tools for 2026, though the extracted text is dominated by unrelated skincare content.

A how-to guide explaining how to design, build, and grow an AI influencer character, with a walkthrough of Pollo AI's influencer and video tools.

Anthropic CEO Dario Amodei predicts AI could turn cancer, Alzheimer's, heart disease, and mental illness into 'relics of the past' within a decade, though pharmaceutical leaders and scientists question whether research can meet that timeline.

An arXiv paper argues that human trajectory prediction models increasingly neglect human behavior and social interaction in favor of quantitative metrics, drawing parallels with interactive art to propose new analytical perspectives.

A sponsored market-research release projects the global AI in drug discovery market will grow from $3.24B in 2026 to $15.54B by 2033 at a 25.1% CAGR, citing pharma adoption, government funding, and cross-industry collaborations.

UNIC Athens hosts the IMCL 2026 conference, featuring 119 papers, four keynotes, and 11 University of Nicosia contributions on how schools and universities use AI in teaching, assessment, and research.

The Financial Action Task Force warns that transnational scam networks are industrializing fraud with AI, deepfakes, and cryptocurrency infrastructure, causing tens of billions in annual losses and requiring faster, end-to-end financial disruption.

A CISO daily briefing warns that attackers are targeting the build and secrets layer of AI applications, including a wfx campaign using 4,837 hijacked GitHub accounts, an agentic pentest framework reaching AWS Secrets Manager, a critical Apache mod_http2 RCE, and a Fireworks-Notion data exposure via a fourth-party AI subprocessor.

Identity verification firm Veriff warns that generative AI is making holiday-season scams cheaper and more convincing through deepfake video, voice cloning, and synthetic identities, sharing shopper tips and calling for AI-based verification.

Tech Digest summarizes Surfshark research tracking 2,430 deepfake incidents across 114 countries and $3.91 billion in fraud losses, with investment scams and social media as primary vectors.

An industry report argues that supply chains must shift from incremental improvement to structural change by prioritizing data discipline, cybersecurity, and network coordination over edge technology deployment.

Naked Capitalism's daily links post aggregates external stories on climate, pandemics, China, the Middle East, and European politics.