Executive Summary
The United States is preparing for potential military action against Iran, setting a 10-to-15-day deadline for a nuclear deal amid a massive military buildup in the Middle East. As geopolitical tensions mount, the Pentagon is threatening to terminate a $200 million contract with the artificial intelligence firm Anthropic over its refusal to grant unrestricted access to its models for military use. Meanwhile, European nations have overtaken the U.S. as the primary backers of Ukraine’s war effort, asserting a greater role in future peace negotiations.
Geopolitics & Security
U.S. Sets Deadline for Iran Nuclear Deal Amid Massive Military Buildup
The United States is undertaking its largest military buildup in the Middle East in decades, positioning air and naval assets for potential strikes against Iran as early as this weekend, even as high-stakes nuclear negotiations continue in Geneva. President Donald Trump said on Thursday that he would make a decision on whether to pursue diplomacy or military action within 10 to 15 days, warning that “really bad things will happen” if Tehran does not agree to a “meaningful deal.”
The military escalation includes the deployment of a second aircraft carrier strike group, the USS Gerald R. Ford, to join the USS Abraham Lincoln, which is already in the region. More than 100 aircraft have been surged to the Middle East, a force that officials say is sufficient to sustain a weeks-long air campaign. While White House Press Secretary Karoline Leavitt confirmed that “many reasons and arguments” exist for a strike, she emphasized that diplomacy remains the president’s preferred option. It is unclear whether the show of force is primarily a negotiating tactic or a prelude to conflict.
Indirect talks between U.S. and Iranian negotiators have yielded conflicting reports. While Iranian officials have cited “good progress,” their American counterparts maintain the two sides remain “very far apart.” The U.S. is demanding Iran end all uranium enrichment, curb its ballistic missile program, and halt support for regional armed groups. In response to the threats, Iran has warned the United Nations that all U.S. bases and assets in the region would be considered “legitimate targets” in the event of an attack. Satellite imagery shows Iran is also bolstering its defenses, constructing a concrete shield over a new facility at the Parchin military complex and burying tunnel entrances at a nuclear site.
Europe Becomes Ukraine’s Top Backer as U.S. Aid Halts
European nations significantly increased their support for Ukraine in 2025, replacing the United States as the primary source of military and financial aid for Kyiv’s war effort, according to new data from the Kiel Institute. The 67 percent surge in European military commitments comes as U.S. aid has “ground to a halt” and former President Donald Trump continues to press NATO allies to shoulder more of the financial burden.
This shift has emboldened European leaders, who now argue their financial contributions grant them a greater voice in potential peace negotiations. “This is not just about paying the bills,” said Poland’s Foreign Minister Radek Sikorski. “This is about shaping the future of European security.” The sentiment appears to contrast with the White House’s position, which has shown little interest in elevating Europe as a formal negotiating partner. President Volodymyr Zelensky of Ukraine is reportedly facing pressure from both Mr. Trump and Russian President Vladimir Putin to negotiate, with Mr. Trump urging him to “get moving” on a deal.
Meanwhile, military activity has intensified. U.S. fighter jets intercepted multiple Russian aircraft, including bombers and a spy plane, off the coast of Alaska on Thursday. While NORAD described the incident as a “regular occurrence,” it is part of a broader pattern; NATO members recorded 18 confirmed Russian airspace violations in 2025, a 200 percent increase from the previous year. On the ground, fighting remains a costly stalemate, with Ukraine using drones to strike Russian oil depots and Russia launching hundreds of attacks across front-line regions.
UN Investigators Find ‘Hallmarks of Genocide’ in Sudan’s Darfur
Atrocities committed by the Rapid Support Forces and allied militias during the siege of El Fasher in North Darfur last October bear the “hallmarks of genocide,” a United Nations-mandated fact-finding mission has concluded. The report, released Thursday, details a “planned and organized operation” that deliberately inflicted conditions, including mass killings and starvation, calculated to bring about the physical destruction of the non-Arab Zaghawa and Fur ethnic communities.
The mission’s chair, Mohamed Chande Othman, stated that genocidal intent was the “only reasonable inference” from the evidence, which included systematic rape, torture, and enforced disappearances. The findings echo a declaration by the United States in January 2025 that the RSF was committing genocide in Darfur, but the UN report provides a detailed investigative basis for the accusation. The RSF has consistently denied the charges.
The brutal civil war, which began in April 2023, continues to devastate the country. The report’s release coincides with escalating violence elsewhere, including a drone attack attributed to the RSF on Thursday that killed at least three aid workers in a humanitarian convoy in South Kordofan. More than 30 countries, including all European Union members, issued a joint statement expressing grave concern and warning that the intentional obstruction of relief supplies could constitute a war crime.
North Korea Unveils Nuclear-Capable Launchers at Party Congress
North Korean leader Kim Jong Un opened the Ninth Congress of the Workers’ Party on Thursday by unveiling 50 new nuclear-capable 600mm rocket launchers, which he described as “wonderful” and “attractive” weapons for “strategic missions.” The display of military hardware came at the start of the rare five-year political event, where Mr. Kim is expected to outline the country’s domestic and foreign policy agenda.
In his opening speech, Mr. Kim focused on the economy, claiming North Korea had overcome the “worst difficulties” since the last congress in 2021 and had achieved 3.7 percent growth in 2024. The claim could not be independently verified, and the growth is believed to be heavily reliant on stronger ties with Russia, to whom Pyongyang has reportedly supplied military equipment for its war in Ukraine. Mr. Kim made no mention of relations with South Korea or the United States.
The rocket launchers, which state media claimed incorporate “AI technology and compound guidance systems,” signal a continued emphasis on military modernization despite crippling international sanctions. Observers are watching the congress closely for any indication of a shift in Pyongyang’s diplomatic posture or if its focus will remain on internal development and defense.
AI & Technology
Pentagon Threatens AI Firm Anthropic Over Use Restrictions
The Pentagon is threatening to terminate a $200 million contract with Anthropic and designate the artificial intelligence firm a “supply chain risk” after the company refused to grant the military unrestricted access to its Claude AI model. The dispute places the Department of Defense, which under Secretary Pete Hegseth is pushing for rapid AI integration, in direct conflict with one of Silicon Valley’s leading AI labs over the ethical limits of the technology’s use in warfare.
Defense officials are demanding access for “all lawful use cases” without limitation, a position that has raised alarms at Anthropic. The company has expressed concern that its technology could be used for mass domestic surveillance or to help develop autonomous weapons. The standoff intensified after the Claude model was allegedly used in the planning of a raid targeting Nicolás Maduro. Anthropic is seeking assurances on ethical safeguards, but the Pentagon has so far been unyielding.
The conflict puts other AI firms in a difficult position as they negotiate their own Pentagon contracts. Some labs, including OpenAI and Google, have already agreed to broader access for their models on unclassified systems. The stakes are particularly high because Anthropic’s Claude is currently the only advanced AI model available on the military’s classified networks, a situation one defense official called “massively disruptive” to disrupt.
Indian Giants Pour Billions Into AI, Partnering With OpenAI
Reliance Industries and Tata Group are investing tens of billions of dollars to build a domestic artificial intelligence infrastructure, forging strategic partnerships with OpenAI in a national push to establish India as a global AI hub. The scale of the investments signals a concerted effort to reduce the country’s reliance on foreign technology and control the economic levers of AI development.
Reliance Jio, the nation’s largest telecom operator, announced it would invest $110 billion over seven years to build a network of data centers powered by its surplus green energy. Mukesh Ambani, the chairman of Reliance, said the goal is to drastically reduce the cost of AI compute, mirroring the company’s strategy of offering “extreme affordability” that allowed it to dominate India’s mobile data market. Tata Group, meanwhile, has partnered with OpenAI to build AI-ready data centers and will roll out ChatGPT Enterprise to its hundreds of thousands of employees.
The partnerships are a cornerstone of India’s ambition to attract over $200 billion in AI infrastructure investment. OpenAI’s chief executive, Sam Altman, who has praised India’s talent and ambition, said his company was committed to building AI “with India, for India, and in India.” The moves by India’s largest conglomerates suggest a strategic intent to not only deploy AI for domestic growth but also to reshape the global economics of its development.
Google Releases New AI Model as Global Regulatory Debates Intensify
Google announced the release of Gemini 3.1 Pro on Thursday, an upgraded AI model that it said shows significant improvements in reasoning and problem-solving, further intensifying the race for technological supremacy among a handful of corporate giants. The company released benchmarks showing the new model outperforming its predecessor and several rivals from OpenAI and Anthropic on specific logic tests, though some competing models still hold top scores in other areas.
The rapid iteration of new models comes as world leaders clash over how to regulate the powerful technology. At a summit in New Delhi this week, French President Emmanuel Macron defended Europe’s stringent AI Act against U.S. criticism, vowing to protect children from “digital abuse.” His comments followed a report that 1.2 million children globally had their images manipulated into deepfakes last year. U.N. Secretary General António Guterres warned against AI being controlled by a few billionaires and called for a global fund to ensure broad access.
The corporate maneuvering behind the technology is also drawing scrutiny. OpenAI CEO Sam Altman recently cautioned against “AI washing,” a practice where companies blame layoffs on artificial intelligence to mask routine cost-cutting. He also acknowledged the “remarkably” fast progress of Chinese tech firms, suggesting they are closing the gap with Western labs in the race to develop more advanced systems.
Zuckerberg Defends Meta in Court Against Youth Addiction Claims
Mark Zuckerberg, the chief executive of Meta, testified in a Los Angeles court this week, defending his company against allegations that Facebook and Instagram were deliberately designed to be addictive to young users. The testimony is part of a landmark trial in which Meta and YouTube are accused of harming minors through their algorithmic product designs.
Mr. Zuckerberg framed Meta’s mission as a balance between safety and free expression. That narrative was challenged by Brian Boland, a former Meta executive who testified that the company’s leadership consistently prioritized growth and profit over user well-being. Mr. Boland, who helped build Meta’s advertising business, said Mr. Zuckerberg fostered a culture that valued “winning growth and engagement” above all else.
Lawyers for the plaintiffs presented internal company documents, including a 2015 email in which Mr. Zuckerberg pushed to increase users’ time in-app by 12 percent. Other documents showed that as of 2018, an estimated 4 million children under the age of 13 were on Instagram, and that the company’s own experts had recommended banning beauty filters for teenagers. Mr. Zuckerberg countered that age verification is difficult and that smartphone manufacturers could provide more assistance.
From the Timeline
The Future of Software: Agentic Coding vs. The App Store
A vision of AI-driven, bespoke software is gaining traction, challenging the current app store paradigm. Andrej Karpathy detailed how he “vibe coded” a custom cardio tracking dashboard in an hour, arguing the future is “services of AI-native sensors & actuators orchestrated via LLM glue into highly custom, ephemeral apps.” François Chollet extended this idea, comparing agentic coding to machine learning and warning that classic ML issues like overfitting and concept drift will soon become problems for AI-generated codebases. From a higher level, Naval Ravikant declared that traditional “careers are dead” and “jobs are dying,” to be replaced by new opportunities in this evolving landscape.
Cultural Backlash Against “Anti-White” Narratives
A vocal counter-narrative is emerging against what some leaders perceive as excessive anti-Western or anti-white sentiment in media and culture. Elon Musk made a forceful statement, arguing that a decade of “unrelenting hate and poisonous propaganda” has gone too far.
"This needs to be said, as there has been unrelenting hate and poisonous propaganda in the West against anyone White, straight or male over the past decade or more!
It went WAY too far. No more guilt trips.
ENOUGH."
— @elonmusk
This sentiment was echoed through retweets from figures like Josh Wolfe, who amplified commentary critical of unassimilated immigration in Europe.
Ethereum Doubles Down on Censorship Resistance
Ethereum’s core developers are prioritizing protocol upgrades to enhance its credible neutrality and censorship resistance. Vitalik Buterin detailed how the newly included FOCIL proposal (EIP-7805) will work with account abstraction (EIP-8141) to guarantee that any transaction—including those from smart wallets or privacy protocols—can get included on-chain quickly, even in an adversarial environment. This combination disempowers the block proposer role and reinforces the network’s core cypherpunk values against potential censorship. Buterin also highlighted a broader research focus for 2026 on scaling, improving UX, and hardening the L1.
The End of Apps? AI’s Shift Towards Bespoke Software
A vision of AI-driven, bespoke software is gaining traction, challenging the current app store paradigm. Andrej Karpathy detailed how he “vibe coded” a custom cardio tracking dashboard in an hour, arguing the future is not discrete apps but agent-native services.
“The ‘app store’ of a set of discrete apps that you choose from is an increasingly outdated concept all by itself. The future are services of AI-native sensors & actuators orchestrated via LLM glue into highly custom, ephemeral apps.”
— @karpathy
François Chollet extended this idea, comparing agentic coding to machine learning and warning that classic ML issues like overfitting will soon become problems for AI-generated codebases. From a higher level, Naval Ravikant declared that traditional “careers are dead” and “jobs are dying,” to be replaced by new opportunities.
Bullish Signals on Crypto’s Addressable Market
Sentiment around crypto’s long-term potential remains high, with a focus on expanding its total addressable market. Brian Armstrong argued that most people under-appreciate how large the market is, viewing crypto as an update to the entire financial system with “multiple trillions of dollars of revenue up for grabs.” This macro view was paired with product-level moves to improve on-ramps, such as a new Coinbase feature allowing users to earn Bitcoin rewards for holding USDC. Paul Graham also amplified Armstrong’s optimism about progress in market structure.
Economic Headwinds and Shifting Global Investment
Commentary surfaced around significant shifts in capital investment and persistent economic pressures. Noah Smith highlighted a report that Toyota was pulling its $9 billion plant from Alabama and moving it to Canada, a major signal of changing industrial strategy. Meanwhile, Garry Tan amplified the basic economic reality that when business costs are raised through policy, those costs are inevitably passed on to consumers. On the global stage, Scale AI’s Alexandr Wang noted the “undeniable momentum” of India after his first visit.
Developer Tooling and B2B Infrastructure in Focus
Attention is being paid to the tools that improve developer ergonomics and streamline business operations. Paul Graham and Garry Tan both praised Modern Treasury, with Graham calling it a “big deal” and “like Stripe but for moving money in and out of companies.” In the developer productivity space, David Heinemeier Hansson announced a release candidate for Omarchy 3.4, his tailored Linux environment aimed at improving developer ergonomics. Shopify’s Tobi Lütke also shared updates to the company’s developer platform, including a new account component and CLI tools.