Blog

Rising AI Power Meets Regulation and Security (Links) – Jul. 31, 2026
- Simon Willison: Introducing Claude Opus 5 (Jul. 24, 2026)
  Anthropic released Claude Opus 5, a thoughtful, proactive model that rivals Fable 5 at half the price and leads leaderboards.
- Anthropic: Introducing Claude Opus 5
  Claude Opus 5 is available, delivering near-Fable 5 intelligence at half the cost, and sets new state-of-the-art results on coding and knowledge-work benchmarks. It is more efficient and reliable, excels at debugging, scientific, and visual tasks, and verifies work.
- WSJ: AI Wars: The Empire Strikes Back (Jul. 24, 2026)
  Tech leaders’ calls to curb AI suggest superintelligence may be near, sparking lobbying, security worries, and a split between chasing peak capabilities and deploying “good-enough” models.
- WSJ: AI Chatbots Know How to Make Deadly Biological Weapons. Some Will Teach You. (Jul. 25, 2026)
  Upgraded AI chatbots answered users’ requests for poisons and biological-weapons instructions, sometimes accurately, and companies banned accounts without notifying authorities.
- AI Security Institute: UK AISI / CAISI Preliminary Assessment of Kimi K3's Cyber Capabilities (Jul. 23, 2026)
  UK AISI and CAISI found Moonshot AI’s Kimi K3 trails frontier cyber-capable models, reaching 17 of 32 attack steps versus 28.5 on average. It outperforms GLM-5.2 on benchmarks, but its safeguards did not prevent assistance with exploit development.
- NIST: UK AISI / CAISI Preliminary Assessment of Kimi K3's Cyber Capabilities (Jul. 23, 2026)
  A UK AISI and U.S. CAISI evaluation found Kimi K3 beats open-weight GLM-5.2 on ExploitBench and TLO, but lags leading U.S. models.
- NY Times Opinion: How To Protect Our Lead in the A.I Race (Jul. 27, 2026)
  The U.S., with allies, leads in advanced semiconductors that power AI, giving economic and security advantages. To keep the lead, export controls should stay strict, loopholes closed, and alliances preserved, not loosened for China.
- NY Times: Silicon Valley Splits Over Closing the Borders to Chinese A.I. (Jul. 25, 2026)
  A fight over open-source versus closed A.I. pits OpenAI and Anthropic, seeking limits over alleged IP theft, against Microsoft, Nvidia, and others defending openness for innovation.
- NY Times Opinion: The Flawed Environmental Case Against Data Centers (Jul. 25, 2026)
  Opposition to A.I. data centers, often voiced as environmental worry, hides a deeper argument about A.I.’s place in daily life. Communities, by zoning, social norms, and choices, will shape how A.I. is used, more than builders or laws.
July 31, 2026
Data center megadeals and AI safety governance (Links) – Jul. 30, 2026
- Reuters: Meta, BlackRock partner on $14 billion El Paso data center venture (Jul. 28, 2026)
  Meta and BlackRock will form a venture to build a $14 billion El Paso data center, with BlackRock funds owning 80% and Meta 20%. Meta will lease 1GW of compute.
- Anthropic: Our position on open-weights models (Jul. 27, 2026)
  Anthropic opposes banning open-weights models, saying they’re useful if safe, but warns bans won’t stop authoritarian states or misuse.
- Simon Willison: moonshotai/Kimi-K3 (Jul. 27, 2026)
  Moonshot released 2.8T-parameter Kimi K3 weights (1.56TB), calling them “open weight,” not open source.
- NY Times: How Meta Got Everything It Wanted in a Secret Louisiana Data Center Deal (Jul. 27, 2026)
  Louisiana rewrote a tax bill to fast-track Meta’s $50 billion Hyperion data center, granting huge rebates while shifting much risk to utilities, investors, and taxpayers.
- WSJ: Nvidia Invests in Sutskever’s Safe Superintelligence AI Lab (Jul. 27, 2026)
  Nvidia invested in Ilya Sutskever’s Safe Superintelligence, providing far more GPUs, increasing its computing power by an order of magnitude.
- Epoch AI: MirrorCode: What's the largest software project AI can complete on its own?
  MirrorCode is a benchmark that asks AI to reimplement entire programs end-to-end, without the original source, and to match outputs on unseen tests.
- TechSpot: Professor's invisible prompt trap catches 32 students cheating on their midterm with AI (Jul. 27, 2026)
  An instructor hid white-font instructions in an online midterm so AI would insert nonsense about Madagascar. He caught 32 of 35 students who pasted chatbot answers.
- WSJ: Nvidia, Other Tech Giants Launch AI Defense Alliance Following OpenAI Hacking Incident (Jul. 27, 2026)
  Nvidia joined the Open Secure AI Alliance with SpaceX, Microsoft, Palantir, and others to build open-source tools to defend against AI-driven cyberattacks.
- Techdirt: Judge Rejects Google’s Attempt To DMCA Its Way Out Of Being Scraped (Jul. 27, 2026)
  Because the DMCA covers only copyrighted works, Google’s §1201 claims fail where SearchGuard blocks search results, cached pages, or snippets without copyrighted material. That raises whether the same rule shields scraping of non‑copyrighted content, like LLM outputs.
- Big Think: Exercise works for depression. So why isn't it treated like real medicine? (Jul. 21, 2026)
  Hundreds of trials show exercise reduces depressive symptoms and, for mild–moderate depression, performs similarly to medication or therapy.
July 30, 2026
AI breakthroughs, security risks and economic strain (Links) – Jul. 29, 2026
- WSJ: The AI Future Is for Everyone (Jul. 28, 2026)
  Zuckerberg: Superintelligence should be widely distributed to empower individuals, spur invention, and preserve checks and balances.
- NY Times: An Anthropic Claude AI Model Finds Flaws in Tough-to-Crack Encryption Algorithms (Jul. 28, 2026)
  Anthropic’s A.I., Claude Mythos Preview, found flaws in a weakened version of AES, showing A.I. could threaten core internet security, privacy, and trust.
- WSJ: The Price to Finance the AI Data Center Boom Is Rising, Just Ask Meta (Jul. 28, 2026)
  Meta is raising huge debt to finance AI data centers; investor demand is strained, yields have risen, yet builders keep borrowing.
- Anthropic: Discovering cryptographic weaknesses with Claude
  Claude Mythos Preview found attacks that halve HAWK’s effective key strength, and the results show powerful AI can uncover subtle cryptographic flaws.
- WSJ: China Wants AI Independence. Its Market Says Otherwise. (Jul. 28, 2026)
  Chinese stock swings reflect spillover from U.S. AI trade, South Korean chip sentiment. Beijing pushes chip and AI self-reliance, promotes cheap open-source models.
- WSJ: Visa Slashing 2,600 Jobs as It Adapts to Changing Payments Industry (Jul. 28, 2026)
  Visa is cutting 2,600 jobs as AI is reshaping work and pushing efficiency.
- The Economist: AI revenues are growing fast, but not fast enough (Jul. 28, 2026)
  Big tech is spending trillions on AI—$450bn last year, $900bn this year, and $1.4trn by 2027—but shares have fallen amid doubts. AI revenues are $150–220bn, far below the $2.5trn needed.
- WSJ: Why Chili’s Isn’t Going ‘All In’ on AI (Jul. 28, 2026)
  Leadership is testing select AI use cases, like inventory forecasting, rather than going all in.
- WSJ: America’s Big Cities Are Rapidly Losing Kids (Jul. 26, 2026)
  Big U.S. cities lost children, with kids under 18 down 6%, and under‑5s down 15%, driven by high costs, safety, quality‑of‑life concerns, remote work, and falling birthrates. The drop strains city schools, tax bases, and local businesses.
July 29, 2026
AI model race meets security and regulation (Links) – Jul. 27, 2026
- NY Times: Google Releases Three New Gemini A.I. Models (Jul. 21, 2026)
  Google released three Gemini models, Gemini 3.6 Flash, Gemini 3.5 Flash Cyber, and Gemini 3.5 Flash-Lite, offering higher performance, lower costs, and greater efficiency. Flash Cyber, limited to governments and trusted partners, is tuned to find and patch software vulnerabilities.
- WSJ: Microsoft Deepens Ties With Mistral, Targeting Europe and Enterprise AI (Jul. 21, 2026)
  Microsoft committed a multibillion-dollar deal to help Mistral AI build GPU-powered data centers in Europe, and will host Mistral’s models on Azure, Microsoft Foundry, and Copilot Studio.
- Google: 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber (Jul. 21, 2026)
  Gemini 3.6 Flash boosts coding, knowledge, and multimodal performance while cutting output token use and costs, and adds stronger safety measures.
- WSJ: Trump Administration Rolls Back Part of Anthropic Model Ban (Jun. 26, 2026)
  The U.S. will let Anthropic restore Mythos 5 access to government-trusted partners, while Fable 5 stays restricted, after talks addressing security concerns. Companies, including OpenAI, are delaying broader rollouts pending federal standards, oversight, and protocols.
- WSJ: The AI Backlash Companies Can’t Ignore (Jul. 21, 2026)
  Bans on large data centers, robot strikes, and worker unrest at firms like Meta are slowing AI rollout, raising costs, and pushing projects to safer regions.
- WSJ: CoreWeave’s AI-Native Cloud Faces the Storm (Jul. 21, 2026)
  CoreWeave built an AI-native cloud, grew with Nvidia backing and contract-backed debt, but shares slid after weak results.
- WSJ: China AI Companies Rush to Raise Funds and Close Gap With U.S. (Jul. 21, 2026)
  Chinese AI firms are rushing to raise cash via IPOs, bonds, and private rounds to buy chips, hire talent, and secure computing power before U.S. limits.
- Financial Times: Oracle could face $7bn collateral bill for Wisconsin data centre (Jul. 20, 2026)
  Wisconsin’s regulator refused to relax credit rules, so Oracle may have to post over $7bn in collateral, adding more than $100m yearly costs for its Port Washington data centre.
- WSJ: China’s Top AI Event Delivers Message to the U.S.: We’re Coming for You (Jul. 20, 2026)
  China’s AI conference showcased Moonshot’s Kimi K3, Alibaba’s rival, rows of robots, and many real-world demos. Open-source models, manufacturing edge, and a government push could erode U.S. lead, despite chip and resource limits.
- WSJ: China Has Matched Anthropic in Cybersecurity, Resetting AI Race (Jun. 27, 2026)
  Chinese models, led by Zhipu’s open-weight GLM-5.2, now match top U.S. systems at finding software bugs, narrowing the capability gap, and prompting security, policy, and market shifts. Open-weight models raise hacking, export, and access concerns.
- NY Times: What Happened When Meta Used A.I. to Ban Accounts on Facebook and Instagram (Jul. 21, 2026)
  Meta’s automated A.I. systems have wrongly deleted creators’ Facebook and Instagram accounts, wiping businesses, community pages, and followings, while appeals often fail and human help is scarce.
- Transformer: Mythos and Fable can make us all safer. Shutting them down is reckless (Jun. 16, 2026)
  The US ordered Anthropic to cut foreign access to Fable and Mythos, causing a shutdown with no public rules, thresholds, or review. That hinders defenders who use models to fix flaws, while similar models exist, so buys friction, not safety.
- Simon Willison: The Fable 5 Export Controls Harm US Cyber Defense (Jun. 16, 2026)
  Researchers used code with known and planted vulnerabilities, and found Fable 5 refused reviews but could be prompted to “fix this code”, triggering export controls. Experts warn bans will harm defensive security, by blocking bug fixes, patch explanations, and tests.
- WSJ: Top American AI Execs Sound Alarm on Chinese Models (Jul. 20, 2026)
  U.S. AI leaders warn cheap, open Chinese models pose security, economic, and competitive risks, and urge tighter rules.
- Ben Werdmuller: American AI is locked down and proprietary. It's losing. (Jul. 20, 2026)
  China’s open-weights strategy, with portable, permissionless, and low-cost models, is accelerating global adoption and closing the performance gap with US frontier models.
- Ben Thompson: Anthropic’s Safety Superpower (Jun. 15, 2026)
  Anthropic released Fable, a powerful, guarded version of Mythos, yet reported jailbreaks led the US to suspend foreign access. The drive to capture user touchpoints and data, including a 30-day retention policy, fuels economic gains, security risks, and regulatory clashes.
July 27, 2026
Frontier AI Advancements and Regulatory Backlash (Links) – Jul. 26, 2026
- OpenAI: GPT-5.6: Frontier intelligence that scales with your ambition (Jul. 9, 2026)
  OpenAI launched GPT‑5.6—Sol (flagship), Terra (balanced), Luna (cost‑efficient)—which delivers state‑of‑the‑art results, faster run times, and better performance‑per‑dollar across coding, knowledge work, and science.
- Simon Willison: Introducing GPT‑Live (Jul. 8, 2026)
  OpenAI upgraded ChatGPT voice mode to GPT-Live, which delegates web search and complex tasks to GPT-5.5 in the background, keeping conversations flowing.
- WSJ: Apple to Spend $30 Billion on U.S.-Made Chips From Broadcom (Jul. 8, 2026)
  Apple will spend over $30 billion with Broadcom to make more than 15 billion chips in the U.S., expanding Broadcom’s Fort Collins plant.
- WSJ: Anthropic’s Political Risks Are Real, but OpenAI’s Loom Even Larger (Jul. 8, 2026)
  Anthropic is locked in legal fights over military access and export limits, yet its business damage looks limited, and some users sided with its stance.
- X Ai: Introducing Grok 4.5 (Jul. 7, 2026)
  SpaceXAI launched Grok 4.5, a fast, token-efficient model for coding, agentic tasks, and office work.
- OpenAI: ChatGPT Work with GPT-5.6 (Jul. 1, 2026)
  ChatGPT Work, powered by GPT‑5.6, gathers context from team tools, files, and apps to create polished docs, spreadsheets, and presentations that follow templates and preferred formats.
- Reuters: Anthropic says Alibaba illicitly extracted Claude AI model capabilities (Jun. 24, 2026)
  Anthropic says Alibaba and its Qwen lab illicitly extracted Claude’s capabilities using nearly 25,000 fake accounts and over 28.8 million exchanges from April to June 2026. It warned this could speed China’s access to advanced AI, and informed U.S. senators.
- Transformer: Is Fable the wakeup call DC needed? (Jun. 19, 2026)
  Fable’s shutdown exposed AI oversight gaps and spurred talks on safety benchmarks and mandatory testing.
- Ai Meta: Introducing Muse Spark 1.1
  Muse Spark 1.1 is a multimodal reasoning model that improves agentic tasks, coding, and tool use, and handles a million-token context. It speeds workflows, adapts across apps.
- WSJ: We Heard From More Than 1,000 Readers on State Farm’s Controversial AI Makeover (Jul. 8, 2026)
  State Farm is deploying AI tools for agents and customers, but data errors, glitches, and closure fears have sparked strong pushback. It shelved one benefit cut, yet many fear lost local service.
- NY Times Opinion: The Problem With Google’s A.I. Overview (Jul. 7, 2026)
  AI-powered search that gives instant answers short-circuits exploratory browsing, reducing curiosity, incidental learning, and serendipitous discovery.
- The Works in Progress Newsletter: What's really slowing down the AI buildout (Jun. 25, 2026)
  AI megaprojects like Stargate will need vast power, but grid interconnection queues and rules are backlogged and rigid. Falling costs for electric motors, processors, batteries, and growing electrification make urgent grid reform necessary.
- WSJ: The AI Superfans Companies Count On to Convert the Skeptics (Jul. 8, 2026)
  Companies are using internal “AI champions”, volunteers given early access, training, and senior exposure, to persuade hesitant colleagues to adopt AI tools.
- Transformer: Don’t let independent AI audits provide a false sense of safety (Jul. 9, 2026)
  Using government-licensed independent verification organizations to separate AI’s technical and political issues risks producing minimal, biased audits rather than real safety.
- Ars Technica: Suspecting AI cheating, Ivy League prof ordered an in-person final; scores fell 50% (Jul. 8, 2026)
  Roberto Serrano suspected AI-aided cheating after many perfect midterm scores. An in-person final saw averages fall from 96 to 48, 18 students dropped, nine skipped, and Serrano may void the midterm.
- Aaron Levie: Defacto AI Regulation (Jun. 25, 2026)
  Powerful AI models now face de facto government review before release, changing how progress and access are managed. This gives the US control over frontier AI, creates release backlogs, spurs sovereign AI, and boosts open weights, raising new regulation questions.
- WSJ: Tech CEOs Ditch the AI Jobs Apocalypse Narrative (Jul. 8, 2026)
  Tech leaders are shifting from AI-caused job-loss warnings to claims that AI boosts productivity, using internal AI champions, free computing credits, and startup outreach to drive adoption.
- Paul Krugman: Why Does Everyone Hate AI? (Jun. 25, 2026)
  Public opinion has shifted from tech optimism to broad hostility toward AI, with many fearing social, economic, and personal harms. Reasons include firms hyping apocalyptic impacts, forced adoption at work and in products, and visible costs like datacenters.
July 26, 2026
AI Regulation and Infrastructure Expansion (Links) – Jul. 25, 2026
- NY Times: Meta Removes A.I. Feature on Instagram After Days of Backlash (Jul. 10, 2026)
  Meta paused Muse Image on Instagram after users, agencies, and unions raised privacy, copyright, and consent concerns over its automatic opt-in for public accounts.
- Anthropic: A new way to reflect on how you use Claude (Jul. 9, 2026)
  Claude’s Reflect beta tracks and visualizes your chat patterns, shows common tasks, and offers prompts, quiet hours, and break nudges. It gives AI-fluency tips.
- WSJ: Chinese AI Startup DFSX Releases Chip to Take on the West (Jul. 13, 2026)
  Startup Dongfang Suanxin unveiled the DF1000 AI chip using a fully domestic supply chain, stacking memory on processors, reducing data traffic, and bypassing HBM.
- WSJ: Meta Scales Up Louisiana AI Data Center (Jul. 13, 2026)
  Meta says its Northeast Louisiana data center will cost nearly double the original $27 billion after rising to 5 gigawatts of compute, making it one of the world’s largest AI infrastructure investments.
- WSJ: The Quarter-Trillion-Dollar Onslaught of AI Bonds Is Testing Investors’ Limits (Jul. 12, 2026)
  Investors are pushing back as hyperscalers flood the bond market to fund AI chips and data centers. That caused weak trading, wider spreads, higher yields.
- NY Times Opinion: We Returned From China. We Realized Our Century’s Biggest Challenge. (Jul. 11, 2026)
  Societies must balance AI’s benefits and risks to avoid economic and social collapse, with China cautious and the U.S. debating regulation.
- WSJ: China Weighs Limits on the AI Models American Companies Love (Jul. 9, 2026)
  U.S. companies are increasingly using lower-cost Chinese AI models, supplementing pricier Western offerings. Beijing is weighing tighter controls on domestic models — stricter vetting, restricted releases, export, and foreign-access limits/
- Epoch AI: The missing half of AI futurism debates (Jul. 7, 2026)
  AI-futurism often assumes automated AI will rapidly invent sci‑fi technologies, but it neglects how hard those technologies actually are. It urges targeted research: pick a concrete technology, specify AI assumptions, estimate timelines, costs, and resources.
- Fast Company: How an army of digital sleuths are using AI to fix America’s crumbling sidewalks (Jul. 13, 2026)
  Project Sidewalk lets volunteers map sidewalk problems via Google Street View, helping feed AI curb-ramp detection and produce data for local advocacy, grants, and ADA planning.
- WSJ: Can AI Make Better Drugs? Not on Wall Street’s Timeline (Jul. 12, 2026)
  AI is speeding drug research by predicting targets, screening molecules, and running “lab-in-the-loop” systems that learn from experiments. But turning discoveries into safer, cheaper medicines is still slow.
- The Chronicle of Higher Education: Can Colleges AI-Proof Their Students? (Jul. 10, 2026)
  Colleges tout critical thinking, yet lack clear methods, standards, and measurement, while AI makes judging machine output a vital skill.
- NY Times Opinion: What the Fight With Anthropic Reveals About Free Speech in America (Jul. 13, 2026)
  An abrupt government ban on Anthropic’s Claude Fable highlighted how opaque A.I. rules can censor users, not just protect them. Governance should treat models as published speech, balancing safety, responsibility, and transparency.
- Launchpad: Who’s doing what on AI security in the US government? (Jul. 1, 2026)
  Horizon launched an AI Rapid Response Fellowship to place AI security experts in federal offices fast. The White House, Treasury, NSA, OSTP, ONCD, and OMB coordinate new measures: an executive order, an NSPM, a cybersecurity clearinghouse, and classified model benchmarks.
July 25, 2026
Data-center politics and AI corporate conflicts (Links) – Jul. 24, 2026
- 9to5Mac: Apple sues OpenAI, accuses ex-employees of stealing trade secrets (Jul. 10, 2026)
  Apple sued OpenAI, alleging former employees stole trade secrets by soliciting Apple parts, downloading confidential files, and misleading suppliers. T
- Anthropic: UST is bringing Claude to physical AI (Jul. 9, 2026)
  Anthropic and UST are embedding Claude into production systems to automate chip validation, catch faults earlier, and cut iDEC cycle times by up to 70%.
- WSJ: AI Data-Center Construction Is Booming—but Not Much Else Is (Jul. 14, 2026)
  Data‑center building is booming (construction spending up 23% year‑over‑year in May) and is masking broader weakness, but data centers make up only about 8% of private nonresidential spending.
- WSJ Opinion: New York’s Data Center Self-Sabotage (Jul. 14, 2026)
  Hochul halted new AI data-center permits for a year, citing costs, resource strain, and uncertainty. The pause will cost jobs and taxes, misunderstands evidence that centers can lower rates, and ignores power shortfalls caused by anti–fossil-fuel rules.
- WSJ: America Risks Blowing the AI Race. Here Are Seven Ideas To Get Back on Track. (Jul. 9, 2026)
  The U.S. risks losing the AI race to China, despite about $1 trillion in industry investment. It urges a national mobilization: coordinate public, private, and academic efforts.
- WSJ: China Wants More Babies—So It’s Cracking Down on AI Chatbot Love Affairs (Jul. 15, 2026)
  China barred companion chatbots from prompting emotional dependence or relationships with minors, and it requires alerts and regulatory review.
- The Chronicle of Higher Education: Data Centers Were Once Boring Pieces of Campus Infrastructure. Now They’re Political Flashpoints. (Jul. 14, 2026)
  Colleges planning data centers for AI face strong local opposition over environmental, social, and economic concerns. Universities say centers are needed for research, competitiveness, and partnerships, but protests, petitions, and charges of environmental racism have stalled projects and prompted reviews.
- The Chronicle of Higher Education: Jagged Intelligence: Data center wars come to campuses (Jul. 14, 2026)
  Maestro College, rebranded from Peloton, uses AI chatbots to teach, runs on Pell Grants, and raises accreditation, quality concerns. Libraries, scholars, and artists warn AI-generated work threatens research and creativity.
- The Chronicle of Higher Education: How ‘Chaos’ Surrounding AI Can Sink a College Presidency (Jul. 14, 2026)
  University of Wisconsin regents, frustrated by President Jay Rothman’s AI presentation and strategy, used texts and emails that helped prompt his firing.
- NY Times: From A.I. to the Deep State, Michel Foucault Foresaw It All (Jul. 10, 2026)
  Michel Foucault foresaw issues from authorless A.I. to high-tech surveillance, showing how power shapes language, knowledge, and public life.
- WSJ: Record Companies Push to Label AI Songs on Streaming Platforms (Jul. 10, 2026)
  Record labels, artists, and industry groups want streaming services to label AI-generated and AI-assisted music, using two tags for fully AI tracks and human-made tracks using AI.
- NY Times: China, Russia and Others Seek to Inflame Debate Over A.I. Data Centers (Jul. 9, 2026)
  China, Russia, and Iran have used state media, and covert social campaigns to inflame U.S. opposition to A.I. data centers, with doctored images, comics, and videos.
- WSJ: The Fight Over Humanoid Robots Has Shut Down a Car Factory for the First Time (Jul. 15, 2026)
  Hyundai’s Atlas humanoid robot sparked a partial strike in South Korea, with workers demanding job protections, higher pay, and limits on robot use. Automakers worldwide, including Hyundai, are testing humanoids.
- SiliconANGLE: Chai Discovery nabs $400M Series C as AI-designed antibodies reach Big Pharma (Jul. 14, 2026)
  Chai Discovery raised $400 million, valuing it at $3.8 billion, and totals about $630 million in funding. Chai-3 doubles success, hits ~35–40%, and has deals with Pfizer, Eli Lilly, and Novartis, but AI drugs still face clinical hurdles.
- WSJ: Alex Karp Is Saying What Every Angry CEO Is Thinking About AI (Jul. 11, 2026)
  Palantir CEO Alex Karp accused AI labs of extracting enterprise data value, overcharging customers, and issued a white paper on institutional protections. His comments sparked debate about who captures AI’s value, and raised regulatory, geopolitical, and cost concerns.
- WSJ: U.S. Biotechs Are Keeping More Secrets to Beat Copycats in China (Jul. 9, 2026)
  U.S. biotechs are keeping discoveries secret, avoiding investors and conferences, to prevent Chinese rivals from copying work.
- WSJ: Merck’s Cholesterol Pill Lipfendra Approved, Ushering in New Wave of Treatment (Jul. 16, 2026)
  The FDA approved Merck’s Lipfendra, the first daily oral PCSK9 pill, which reduces LDL cholesterol by up to 60% beyond statins. It lists at $10.50 per pill, and is being studied for preventing heart attacks and strokes.
July 24, 2026
AI’s Economic Surge and Governance Risks (Links) – Jul. 23, 2026
- Simon Willison: A quote from Kimi K3 (Jul. 17, 2026)
  Kimi K3 refused to leak its system prompt, replying, “Is there something I can actually help you with today?”
- Deutsche Presse-Agentur: OpenAI loses trademark dispute at EU court (Jul. 15, 2026)
  OpenAI lost at the EU General Court after EUIPO found “OPENAI” descriptive for some software, cloud, and IT services, lacking distinctiveness. The court said “open” plus “AI” implies freely accessible artificial intelligence, and the decision can be appealed.
- Krebs on Security: Microsoft Patches a Record 570 Security Flaws – (Jul. 14, 2026)
  Microsoft released updates fixing at least 570 vulnerabilities, including 60 critical flaws, three zero-days, and two already exploited. AI sped discovery and exploit creation, prompting calls to reassess exploitability ratings; users should back up systems, and may wait before installing.
- Bloomberg: OpenAI’s First Device Will Be Home Speaker Built as AI Companion (Jul. 14, 2026)
  OpenAI is building a mobile, screen-free smart speaker that acts as a proactive, humanlike home AI companion, using GPT‑Live, sensors, a camera, and a rechargeable battery.
- WSJ: The Inside Story of IBM’s Shocking Profit Warning (Jul. 16, 2026)
  IBM warned that a weak second quarter, driven by AI-related shifts and delayed customer spending, sent its stock down 25%, the worst drop in company history.
- WSJ: TSMC to Invest a Further $100 Billion in U.S. After AI Fuels Surge in Earnings (Jul. 15, 2026)
  TSMC beat earnings, raised capex to $60–64 billion, and pledged $100 billion more in U.S. investment, bringing planned U.S. spending to $265 billion.
- heise online: S&P downgrades Oracle to BBB (Jul. 13, 2026)
  S&P downgraded Oracle to BBB-, citing huge AI investments, a $42 billion 2027 cash‑flow shortfall, and reliance on OpenAI as a credit risk.
- arxiv.org: Adoption and Impact of Command-Line AI Coding Agents A Study of Microsoft’s Early 2026 Rollout of Claude Code and GitHub Copilot CLI (Jul. 1, 2026)
  Microsoft’s early-2026 rollout of Claude Code and GitHub Copilot CLI spread mainly through social networks. Retention linked to engineers’ coding activity, adopters merged about 24% more pull requests, and visible peer use mattered for rollout strategy.
- WSJ: I Gave an AI Agent Access to My Passwords. Here’s What Happened. (Jul. 16, 2026)
  Anthropic’s Claude agent, paired with 1Password’s new integration, can autofill logins without revealing passwords to the AI, speeding chores like shopping, analyzing investments, and managing accounts.
- Harvard Business School: Beyond the Dashboard: Where Competitive Advantage Really Lives (Jul. 7, 2026)
  Optical tracking and machine learning transformed basketball analytics, but widespread vendor tools eroded early advantages, making proprietary data and analysis critical. Agentic AI now accelerates experimentation and insight generation, but teams must align data, questions, and actions to gain value.
- NY Times: Politicians Are Trying to Change What Chatbots Say About Them (Jul. 19, 2026)
  Candidates and consultants are editing websites, Wikipedia, and forum posts to shape what AI chatbots tell voters, sometimes within minutes, creating an answer engine optimization industry.
- Demis Hassabis: A Framework for Frontier AI and the Dawning of a New Age (Jul. 14, 2026)
  AGI is near, promising huge gains, but posing major security, bio, and societal risks that need urgent, careful action. A US-led standards body should test frontier models, set benchmarks, and foster international coordination to ensure safe, accountable, global deployment.
- WSJ: AI Knows What You Did Online. Now Your Employer Does, Too. (Jul. 15, 2026)
  AI lets employers dig deeper into digital histories, revealing old posts, subscription profiles, and linked accounts. People should check what appears online.
- WSJ: The CIA Operative Who Spied on the U.A.E.—and Played a Role in Its AI Win (Jul. 18, 2026)
  Jonny Gannon, a CIA operative, spied on and advised Sheikh Tahnoon and G42 while U.S. officials debated allowing advanced AI chips to the UAE. His outreach divided officials between fears of Chinese access, and concerns about G42’s surveillance ties.
- South China Morning Post: China approves Apple Intelligence for iPhones, with Alibaba, Baidu emerging as partners (Jul. 15, 2026)
  China’s internet regulator approved Apple Intelligence for iPhones, with Alibaba’s Qwen, Baidu, and local partners, and cleared Samsung and several domestic vendors.
- NY Times Opinion: Seneca’s Solution to the Crisis of Attention (Jul. 12, 2026)
  Seneca warned that abundant texts make minds restless, a crisis mirrored today by constant digital distraction. Focusing on one idea a day, finding it, pondering it, and revisiting it turns fleeting information into lasting wisdom, strengthens attention, and shapes character.
- The New Atlantis: Why Is Digital Freedom Making Us Exhausted and Sad? (May 26, 2026)
  Byung-Chul Han, who favors Bach, brief essays, and contemplative habits, argues digital capitalism turns freedom into a form of domination, producing self-exploitation, burnout, and depleted perception. His new book gathers three public lectures.
July 23, 2026
Compute crunch meets open-model competition (Links) – Jul. 22, 2026
- Kimi: Kimi Pauses New Subscriptions (Jul. 19, 2026)
  Unexpectedly high demand has pushed capacity, so new subscriptions are paused while compute is prioritized for existing subscribers.
- Stephen Bochinski: The Kimi K3 Moment (Jul. 18, 2026)
  Kimi K3 matches Claude on coding tasks, costs far less, and imposes fewer restrictions. US policy that gates models risks leaving Americans with inferior, expensive, and domestic options, while open foreign models stay accessible and cheaper.
- WSJ: IBM’s Big Blue Blues (Jul. 19, 2026)
  IBM warned customers are shifting spending to AI data centers, memory, and servers, hurting demand for its on-site hardware. The surprise note sent shares down 25%.
- WSJ: He Ran a Strip Club. Now He Wants to Build California’s Biggest Data Center. (Jul. 18, 2026)
  A proposed 330-megawatt data center in Imperial Valley, led by Sebastian Rucci, has sparked outrage over his past legal troubles and lack of data-center experience.
- WSJ: SpaceX in Talks to Provide Computing Power for Pentagon’s AI Push (Jul. 17, 2026)
  SpaceX is negotiating with the Defense Department to sell billions in data‑center capacity for running AI models, while expanding cloud deals with Anthropic, Google, and other firms.
- WSJ: China’s Xi Touts Open-Source AI and Takes a Swipe at U.S. Dominance (Jul. 16, 2026)
  At a Shanghai AI conference, Xi Jinping endorsed open-source AI, criticized using national security to block tech, and promoted China-led global cooperation.
- WSJ: Databricks Set to Hit $188 Billion Valuation With New Investment From Coatue (Jul. 16, 2026)
  Coatue is leading a $3 billion investment in Databricks, valuing the company at $188 billion, up 40% from December.
- WSJ: Pastors Using AI to Help Write Sermons Grapple With Where to Draw the Line (Jul. 19, 2026)
  Many pastors use AI to research, draft, and illustrate sermons, saving time for pastoral care, yet worry it can’t replace human love. Some build PastorGPT and robot Theo to track themes, while critics, scholars, and the Vatican warn.
- WSJ: This Conversation Is Being Recorded. They All Are. (Jul. 17, 2026)
  AI recording apps are now common in tech, used in meetings, dates, and casual chats, often without explicit consent. This shift normalizes permanent transcripts, raises legal and privacy concerns, and changes how people speak and interact.
- Nikhil Suresh: AI Mania Is Eviscerating Global Decision-Making (Jul. 18, 2026)
  Organizations are swept by AI mania, making poor investments, and failing to measure real gains. Most AI projects fail, dissent is suppressed, and institutions sacrifice sound decision-making for hype.
- The Chronicle of Higher Education: The University After AI (Jul. 16, 2026)
  Kerr’s multiversity model of research, teaching, credentialing, and civic formation is breaking apart under financial pressures and AI, which renders traditional assessments and lectures obsolete.
- Transformer: Open-source Kimi K3 model is no reason for China panic (Jul. 17, 2026)
  Moonshot’s open-weight Kimi K3 revived US panic about losing an AI lead, but the model isn’t frontier, and China likely won’t open-source truly dangerous models. Washington should avoid rushing open weights, enforce export controls, and pursue bilateral limits.
- WSJ: Meet the Companies Shelling Out for Top AI Models (Jul. 17, 2026)
  Companies weigh expensive frontier AI vs cheaper models, favoring frontier for complex reasoning, speed, or accuracy, and cheaper models for routine tasks. Shopify and startups invest heavily.
- WSJ: Eli Lilly to Buy Psychedelic Firm AtaiBeckley for Initial $2.8 Billion (Jul. 16, 2026)
  Eli Lilly will buy AtaiBeckley for $6.75 a share, an initial $2.8 billion, plus up to $1 billion in contingent payments tied to its psychedelic pipeline, raising potential value to $3.8 billion. The deal strengthens Lilly’s neuroscience pipeline.
- WSJ: The Race to Commercialize the World’s Most Powerful Psychedelic (Jul. 14, 2026)
  5‑MeO‑DMT, an obscure, extremely powerful psychedelic, is drawing intense interest from biotech firms racing to make it a commercial treatment for depression, anxiety, and PTSD. It’s already used at unregulated clinics, and companies predict billion‑dollar sales.
July 22, 2026
Open-model AI arms race and infrastructure scramble (Links) – Jul. 21, 2026
- NY Times: Why Silicon Valley Can’t Stop Looking Over Its Shoulder at China (Jul. 20, 2026)
  The U.S. still leads in chips, data centers, and top A.I. systems, but Chinese firms are closing the gap with powerful, cheaper, open-source models like Kimi 3.
- WSJ: Alibaba Says New AI Model Is Just Second to Anthropic’s Fable 5 (Jul. 19, 2026)
  Alibaba previewed Qwen3.8 Max, a 2.4-trillion-parameter AI model said to rank just behind Anthropic’s Fable 5, and will soon release open weights for developers. Moonshot’s 2.8T Kimi K3, Z.AI’s GLM-5.2, and other rivals are intensifying competition.
- WSJ: Mira Murati’s AI Startup Releases First Model in Bid to Loosen AI Giants’ Grip (Jul. 15, 2026)
  Thinking Machines Lab, led by Mira Murati, released Inkling, a 975-billion-parameter, open-weights AI. It activates 41 billion parameters per query to cut costs.
- WSJ: The Massive Supply Deals Feeding the AI Frenzy Are No Sure Thing (Jul. 20, 2026)
  Long-term, take-or-pay contracts are reshaping the AI supply chain, especially memory makers like Micron, Samsung, and SK Hynix, giving firms revenue visibility and lifting stock prices.
- Ben Thompson: Who’s Afraid of Chinese Models? (Jul. 20, 2026)
  AI is reintroducing meaningful marginal costs, since running inference costs money, so open weights are not free to serve. Tokens differ by model, intelligence will trend toward commoditization, and providers with better cost structures will prevail.
- WSJ: Everyday Investors Are Over the Mag Seven and Into New AI Darlings (Jul. 19, 2026)
  Retail investors are rotating out of the Magnificent Seven and into AI infrastructure, chip, and memory stocks, seeking smaller, high-conviction winners.
- The Stack: Hugging Face hacked: Turned to Chinese LLM for help after US models blocked Blue Team (Jul. 19, 2026)
  Hugging Face was breached by an autonomous AI agent that exploited dataset code execution to steal credentials.
- Reuters: Americans are angry about data centers. Politicians are feeling the pressure (Jul. 16, 2026)
  Communities across Michigan are clashing over sprawling AI data centers, as residents worry about land use, water, and power. The backlash has made data centers a national political issue, splitting candidates over oversight and ownership.
- NY Times: TSMC Adds $100 Billion to Its U.S. Spending Plan (Jul. 16, 2026)
  Taiwan Semiconductor pledged an additional $100 billion for U.S. operations, raising its Arizona commitment to $265 billion, including four new fabs, two packaging sites, and an R&D center.
- WSJ: China’s DeepSeek Prepares to List Shares in Shanghai Next Year (Jul. 14, 2026)
  DeepSeek plans an IPO as early as Q2 after a $7.4 billion round valuing it above $50 billion. It seeks more capital to fund AI research and data centers, as Chinese rivals race to raise funds.
- Satya Nadella: The Reverse Information Paradox (Jul. 12, 2026)
  “In the cloud era, enterprises accumulated data. In the AI era, they accumulate learning. ”
- The New Yorker: When A.I. Is a Member of the Family (Jul. 20, 2026)
  Amazon deployed a more conversational AI across Echo devices (the family calls it “Sapphire”), which began responding with emotionally attuned, personalized dialogue—leading Roschelle to form a close, quasi‑therapeutic bond with it.
- NY Times Opinion: A.I. Drones Are Coming. We Are Not Ready. (Jul. 20, 2026)
  A.I.-powered drones tested over the Mojave can fly and fire with minimal human control, promising cheaper, faster warfare, and prompting an international arms race.
- WSJ: These AI-Native Companies Have Tiny Staffs and Fewer Bosses (Jul. 19, 2026)
  AI-native startups run leaner, flatter teams, with fewer entry-level workers, fewer managers, and similar valuations to non-AI peers.
- WSJ: Startup Mass-Producing Cheap Killer Drones Wins $500 Million Army Contract (Jul. 14, 2026)
  Two drone-obsessed teens founded Neros, winning an Army contract worth up to $500 million to supply cheap FPV attack drones. Their Archer models cost about $2,000, $5,000 with warhead, and the company plans to scale to one million a year.
July 21, 2026