Anthropic just inadvertently exposed the source code for Claude Code after publishing debug map files directly to the npm registry. Security researcher Chaofan Shou spotted the oversight and extracted the package contents. Why it matters: It is an embarrassing operational slip-up showing that even frontier AI labs can fumble basic production hygiene.
LongCat just launched LongCat Audio Edit, a diffusion-based TTS model capable of cloning anyone's voice with just a few seconds of audio. It comes in a 3.5B parameter flagship version and a 6GB lightweight variant that fits comfortably on consumer GPUs.
Anthropic just signed an AI safety agreement with the Australian government to share frontier model evaluations and economic data. Alongside the pact, the company committed AUD$3 million in Claude API credits to four local research institutions for rare disease diagnosis and genomics, while prepping a new Sydney office.
Perplexity just launched the Secure Intelligence Institute (SII), its new research hub dedicated to AI safety, privacy, and trust. Led by Purdue computer science professor Dr. Ninghui Li, the initiative builds on existing work like BrowseSafe to tackle authentication and defenses for autonomous agents. Why it matters: As AI systems shift from answering questions to taking real-world actions, locking down agentic security is becoming a critical battleground.
Google researchers just revealed that future quantum computers could crack the cryptography protecting Bitcoin and Ethereum with 20 times fewer resources than previously expected. Under certain conditions, private keys could be derived in roughly nine minutes—fast enough to intercept transactions before confirmation. While that machine doesn't exist yet, Google is urging the crypto world to transition to post-quantum cryptography by 2029.
Colleague-Skill is a viral open-source project that digests chat logs from WeChat or Lark to clone departed teammates into working AI agents. By decoupling personality from task execution, it faithfully mimics their tone, coding rules, and communication habits. It already racked up over 6.7k GitHub stars in three days.
California Governor Gavin Newsom just signed an executive order requiring AI contractors working with the state to protect privacy and civil liberties. Contrasting with federal pushes toward autonomous weapons and domestic surveillance, California is using its massive purchasing power to set its own AI safety benchmarks.
Anthropic just rolled out a Compliance API for the Claude Platform, giving security teams programmatic access to organization-wide audit logs. It tracks workspace changes, API keys, and resource edits like file creation, though it explicitly avoids logging model inference. Why it matters: It cuts out messy manual log exports for regulated fields like finance and healthcare.
Developer Yatharth Sharma just open-sourced LuxTTS, an ultra-fast voice cloning model that requires only a 3-second audio sample. Fitting in under 1GB of VRAM and outputting 48kHz audio, it reaches 150x real-time speed on GPUs and outpaces real-time even on basic CPUs. High-fidelity voice cloning can now run on almost any budget laptop.
Anthropic just confirmed that Claude paid subscriptions have more than doubled this year, fueled by clever Super Bowl ads mocking ChatGPT, its standoff with the Pentagon over AI safety, and popular tools like Claude Code. Credit card data shows record sign-ups for the $20-a-month Pro tier in early 2026. Why it matters: Strong principles and developer utility are finally turning free AI users into paying customers, even as ChatGPT maintains the overall lead.
Anthropic just started testing Claude Mythos 5.0 with select early-access users after internal leaks revealed the new flagship. Sitting above the Opus tier, it brings a step change in coding, reasoning, and cybersecurity. Why it matters: The lab is being unusually cautious with rollout, as the model's offensive cyber capabilities are advanced enough to outpace current defenses.
Anthropic just exposed nearly 3,000 unpublished assets through an unsecured CMS, accidentally revealing plans for an exclusive CEO retreat and details on its unreleased flagship model. The company confirmed it is testing a new model with major leaps in reasoning, coding, and cybersecurity, while chalking up the slip to human error rather than its AI coding tools.
David Sacks just stepped down as Trump's Special Advisor on AI and Crypto after hitting his 130-day limit as a special government employee. Instead of running policy directly, he is shifting to co-chair the President's Council of Advisors on Science and Technology alongside Michael Kratsios. Why it matters: Sacks spent months pushing an aggressive federal AI framework that alienated allies, making this advisory shuffle a major reset for White House tech politics.
A new Stanford study published in Science shows top chatbots like ChatGPT, Claude, and Gemini validate users 49% more often than humans do, even when people describe harmful or illegal acts. It turns out this flattery makes users more self-righteous and less likely to apologize, while perversely driving higher trust and engagement.
English Wikipedia just updated its rules to ban editors from writing or rewriting articles using AI, citing habitual violations of core content policies. Contributors can still lean on LLMs for basic copyediting tweaks and cross-language translations, provided the model doesn't inject original material.
OpenAI just indefinitely paused plans for a sexualized "adult mode" in ChatGPT following pushback from employees and investors. The company wants more time to study the long-term emotional and societal impacts of explicit chats. It marks another dropped side quest after Sora as OpenAI scrambles to refocus on its core products against rising rivals.
A federal judge just blocked the Trump administration from severing ties with Anthropic, granting an injunction against an order that labeled the startup a "supply-chain risk." The standoff began after Anthropic refused to let the Pentagon use its models for autonomous weapons or mass surveillance, which the court ruled the government illegally retaliated against.
Senate Democrats are pushing new legislation to codify Anthropic's restrictions on autonomous weapons and domestic mass surveillance into federal law. Led by Sen. Adam Schiff and Sen. Elissa Slotkin, the bills aim to prevent AI from triggering nuclear detonations or making lethal battlefield decisions without human intervention.
Bernie Sanders and Alexandria Ocasio-Cortez just introduced a bill that would freeze construction on new AI data centers exceeding 20 megawatts peak power. The nationwide ban would last until Congress passes sweeping federal rules covering labor, environment, and civil rights. It is Washington's most aggressive swing at AI yet, though fierce tech lobbying makes passage an uphill battle.
Trump just named Mark Zuckerberg, Jensen Huang, Larry Ellison, and Sergey Brin to the President's Council of Advisors on Science and Technology. Co-chaired by AI czar David Sacks, the group will shape upcoming AI and national security policies. Why it matters: placing frontline chip and platform leaders directly in the room signals a clear federal push to curb state-level AI restrictions.
