Executive Summary
U.S. envoys are mediating tense peace talks between Russia and Ukraine and separate nuclear negotiations with Iran, even as military posturing and deep mistrust complicate diplomatic efforts. In Europe, regulators have launched sweeping investigations into X, Meta, and TikTok over the spread of AI-generated harmful content and potential data privacy violations. Meanwhile, a new report details sustained infiltration of U.S. energy networks by Chinese state-sponsored hackers, and a major U.S.-Japan investment deal highlights a transactional approach to foreign policy under the Trump administration.
AI & Technology
Europe Launches Sweeping Probes of X, Meta, and TikTok Over AI Harms
European regulators have opened a broad, multi-front inquiry into some of the world’s largest social media companies, scrutinizing Elon Musk’s platform X, its AI chatbot Grok, Meta, and TikTok for allegedly spreading AI-generated harmful content and violating data protection laws. The actions represent one of the most significant regulatory challenges to date against the integration of generative artificial intelligence into major online platforms.
In Ireland, the Data Protection Commission, which serves as the lead E.U. supervisor for many U.S. tech firms, launched a large-scale inquiry into X and Grok. The probe will examine potential breaches of the General Data Protection Regulation (GDPR) related to the creation and publication of non-consensual, intimate, or sexualized images, including those depicting children. This follows a wider European Commission investigation into xAI, Grok’s developer, under the Digital Services Act (DSA) for its role in disseminating such content. The Spanish government has also ordered its prosecutors to investigate X, Meta, and TikTok for allegedly spreading AI-generated child sexual abuse material, with Prime Minister Pedro Sanchez stating the platforms are “undermining the mental health, dignity, and rights of our children.”
The regulatory pressure has been matched by direct law enforcement action. French investigators raided the Paris offices of X in early February as part of an inquiry into the platform’s algorithms and AI-generated abuse material. French prosecutors have since summoned Mr. Musk and X’s chief executive, Linda Yaccarino, for interviews. The company has called the raid an “abusive act of law enforcement theatre” and maintains that it removes child sexual abuse material and non-consensual nudity. Last month, X said it implemented technical measures to limit Grok’s ability to generate certain explicit images, but critics, including the Center for Countering Digital Hate, have reported that millions of sexualized images have been created using the tool.
The investigations will assess whether the companies adhered to core E.U. legal principles, including lawful data processing, privacy by design, and conducting risk assessments for new technologies. The outcome could lead to substantial penalties, with fines reaching up to 4 percent of a company’s global revenue under GDPR and 6 percent under the DSA. The coordinated actions by regulators in Ireland, Spain, France, and at the E.U. level are poised to set a precedent for how AI tools are governed on social media platforms across the continent and potentially worldwide.
Chinese Hackers Infiltrate U.S. Energy Grids, Report Finds
Chinese state-sponsored hacking groups have successfully embedded malicious software deep within United States critical energy infrastructure, according to a new report from the cybersecurity firm Dragos. The activity, attributed to a group identified as Voltzite and linked to the well-known Volt Typhoon crew, appears focused on gaining long-term, persistent access to operational control systems, suggesting a strategic aim to be able to disrupt energy distribution at a future date rather than to conduct espionage or steal intellectual property.
The sustained infiltration of U.S. energy networks by a foreign adversary poses a direct threat to national security. The focus on operational technology—the systems that manage the physical processes of power generation and distribution—indicates an intent to prepare for potential sabotage during a future conflict or geopolitical crisis.
Separately, unconfirmed claims of a major security breach have surfaced regarding Palantir Technologies, a data analytics firm with deep ties to U.S. and allied intelligence agencies and military organizations. The allegations were made by Kim Dotcom, an internet entrepreneur with a history of clashing with U.S. authorities. The veracity of the claim could not be independently verified, and Palantir has not commented on the matter. If true, a breach of Palantir’s systems could represent a significant compromise of sensitive government and corporate data.
These developments in state-sponsored and alleged corporate espionage come as new research calls into question the security of popular password management services. A technical analysis suggests that the “zero knowledge” architecture advertised by companies like Bitwarden, Dashlane, and LastPass—a design meant to prevent the companies themselves from accessing user data—may have weaknesses. The research indicates that a server compromise, particularly when features like account recovery or group sharing are enabled, could potentially expose users’ encrypted password vaults.
Geopolitics & Security
U.S. Mediates Ukraine Peace Talks as Zelensky Resists Trump’s Pressure
Delegations from Russia and Ukraine convened in Geneva this week for a third round of U.S.-brokered peace talks, with an American envoy claiming “meaningful progress” even as Ukrainian President Volodymyr Zelensky publicly pushed back against pressure from former President Donald J. Trump to make concessions. The talks opened against a backdrop of continued Russian aerial assaults, with Ukrainian officials reporting that nearly 30 missiles and over 400 drones had targeted 12 regions just hours before the delegations met.
Steve Witkoff, Mr. Trump’s special representative, stated after the initial meetings that both sides had agreed to update their leaders and continue working toward a deal. However, Mr. Zelensky expressed frustration with Mr. Trump’s public posture. In an interview with Axios, he said the former president’s calls for Ukraine to “come to the table fast” and make concessions were “not fair,” adding that any deal requiring the surrender of territory would be rejected by the Ukrainian public. The core disagreement remains control over the Donbas region, where Ukraine still holds about 10 percent of the territory.
The negotiations, which involved separate meetings between Ukrainian negotiators and representatives from the U.S. and European allies, focused on “practical issues and the mechanics of possible solutions,” according to Rustem Umerov, Ukraine’s lead negotiator. A source close to the Russian delegation described the atmosphere as “very tense.” While Mr. Zelensky has stated his readiness to refrain from strikes and seek a “worthy agreement,” the ongoing Russian attacks on Ukrainian cities, which have left hundreds of apartment blocks without heat and electricity in subzero temperatures, highlight the immense gap between the diplomatic efforts and the reality on the ground.
Iran and U.S. Report Progress in Nuclear Talks Amid Military Buildup
The United States and Iran have reached an agreement on “general guiding principles” for a potential nuclear deal following a round of indirect talks in Geneva, though the diplomatic progress was shadowed by a significant U.S. military buildup in the Persian Gulf. The discussions, facilitated by Omani intermediaries, concluded with both sides reporting a path forward, but American officials cautioned that many difficult details remain unresolved.
Iran’s Foreign Minister, Abbas Araghchi, said the talks were “serious, constructive and positive” and that a “clearer path ahead” had emerged. A U.S. official, who spoke on the condition of anonymity to discuss the sensitive negotiations, confirmed that progress was made but noted that “a lot of details” still need to be worked out. Iran is expected to return with detailed proposals within two weeks to address what the official called “open gaps.” The U.S. delegation reportedly included Jared Kushner and Steve Witkoff.
In a novel diplomatic tactic, Tehran has also framed a potential deal as a “trillion-dollar opportunity” for American businesses, highlighting access to its vast energy and mineral reserves in an apparent appeal to President Trump’s focus on economic benefits. Despite Mr. Trump’s expressed optimism that Iran wants a deal, the talks are proceeding amid deep mistrust. Iran’s Supreme Leader, Ayatollah Ali Khamenei, rebuked U.S. rhetoric on regime change and issued a veiled threat, suggesting U.S. warships in the region could become targets. The U.S. has reportedly moved a second aircraft carrier strike group and over 50 fighter jets to the region, signaling a dual-track approach of diplomacy backed by a show of force.
European Nations Accelerate Defense Spending in Response to Russian Threat
European nations are moving to significantly increase their defense budgets and overhaul military procurement, driven by Russia’s continued aggression and a perceived decline in the reliability of U.S. security guarantees. The European Union has established a new funding mechanism, Security Action for Europe (SAFE), with up to $178 billion available to bolster continental arms production and joint procurement.
Germany is at the forefront of the buildup, with plans to invest approximately $77 billion over the next five years, a sum that could make its defense budget the world’s third largest by 2030. The United Kingdom is also considering accelerating its plan to raise defense spending to 3 percent of its gross domestic product. These national efforts are complemented by joint European projects to develop long-range and hypersonic weapons.
The push for greater “strategic autonomy,” however, has exposed friction among key allies. German officials have publicly criticized French President Emmanuel Macron for championing European defense independence while, in their view, failing to sufficiently increase Paris’s own military budget. Berlin has gone so far as to suggest that France should cut social spending to fund its military. Russia has dismissed the European military buildup as “fearmongering” and a threat to its own security. The rearmament drive comes as Russian officials and analysts have begun to publicly question the effectiveness of the United Kingdom’s nuclear deterrent in protecting the rest of Europe, adding to a climate of heightened nuclear anxiety.
Trump’s Transactional Diplomacy on Display in U.S.-Japan Investment Deal
The Trump administration’s transactional approach to foreign policy was highlighted by a new $36 billion investment deal with Japan focused on American oil, gas, and critical minerals. The agreement, part of a broader $550 billion Japanese commitment, was framed by officials as a strategic move to strengthen bilateral ties and counter China’s economic influence in key resource sectors.
President Trump has emphasized the role of tariffs in securing such agreements, reflecting a strategy that often blends economic pressure with diplomatic negotiation. This approach is being closely watched by allies and adversaries alike. Some analysts suggest it creates a “good cop/bad cop” dynamic, where aggressive stances are used to create leverage before offers of reconciliation are made to extract concessions. While potentially effective in specific negotiations, critics worry the method risks alienating traditional allies and creating long-term instability.
The focus on bilateral, transactional deals comes as U.S. intelligence agencies are recalibrating their strategies for a more contested global order. The shift is marked by accelerating technological competition and what some see as a declining faith in international norms and institutions. At the same time, concerns have been raised by policy analysts about the potential for a Trump administration to blur the lines between military and civilian roles, citing the use of military forces for domestic law enforcement in some Latin American countries as a cautionary tale for democratic institutions.
From the Timeline
The Great AI Capex Debate: Build vs. Buy
A strategic divergence in AI is becoming apparent, with @chamath outlining a framework where Big Tech (Amazon, Microsoft, Google, Meta) is investing heavily to own power, compute, and algorithms, betting that this creates insurmountable leverage. He contrasts this with Apple’s capex cuts, framing their strategy as a bet that AI is a feature to be licensed later, a rational but potentially risky assumption that they can rely on their distribution moat.
AI Model Wars Shift to Cost and Practicality
The focus in AI model evaluation is shifting from abstract benchmarks to real-world utility and cost-effectiveness. David Heinemeier Hansson championed Kimi K2.5, noting it is “hilariously cheap” and has become his primary coding assistant over more expensive models like Claude Opus. This aligns with a sentiment from @chamath that side-by-side comparisons on controversial prompts are more useful than “evals that consumers have zero idea how to interpret.”
Coinbase Mounts Defense Against Wall Street Skepticism
Coinbase CEO Brian Armstrong pushed back against what he describes as Wall Street’s misunderstanding of his company, framing it as a classic innovator’s dilemma. He argued that while half of traditional finance is leaning in, the other half resists crypto due to threatened incentives, comparing them to “horse and buggy makers” looking at the automobile. Armstrong also highlighted emerging regulatory clarity, pointing to the CFTC’s exclusive jurisdiction over prediction markets as a clear positive signal.
The Unseen Bottleneck in AI Research: GPU Scheduling
John Carmack surfaced a critical infrastructure challenge below the level of hyperscale data centers, focusing on the inefficiency of GPU scheduling for research. He detailed the need for a clean way to preempt long-running GPU tasks to prioritize more urgent jobs, a feature that is technically feasible but not an incentive for hardware makers to build.
“I wish there were a clean way to preempt GPU tasks, so long running tasks could be transparently paused to allow higher priority tasks to get the minimum time-to-results… Of course, Nvidia would prefer you to Just Buy More GPUs!”
— @ID_AA_Carmack
Major Venture Capital Personnel and Funding Shifts
Significant moves in the venture world drew commentary, with @sama celebrating his brother Jack Altman’s move to Benchmark as an “absolute triumph for the firm.” Separately, Altman also praised Thrive Capital’s new $10B fund, commending partner @JoshuaKushner and the firm’s concentrated, high-conviction investment strategy.
AI-Powered Developer Tooling Matures
Commentary highlighted the growing maturity of AI-native developer workflows, moving from novelty to practical application. @chamath shared a detailed testimonial from a business owner who, despite not having coded since the 1990s, used the 8090 platform to rapidly prototype a mobile app concept. This practical application was echoed by @dhh, who shared his enthusiasm for programming his own developer environment to integrate Neovim, Opencode, and a terminal.
Tech Leaders Amplify Scrutiny of Political and Financial Figures
Thought leaders continued to use their platforms to amplify specific political narratives and scrutinize public figures. @DavidSacks repeatedly shared posts investigating Reid Hoffman’s stated interactions with Jeffrey Epstein. Elsewhere, voices like @garrytan and @ylecun amplified content critical of domestic tax policy and government spending, respectively.