Anthropic
Anthropic business and news from across the web.- Claude Code leak suggests Anthropic is working on a 'Proactive' mode for its coding toolA source code leak from Anthropic's Claude Code tool has revealed potential upcoming features, including a 'Proactive mode' that allows the AI to work without user prompts and a crypto-based payment system. The leak occurred due to a packaging error, not a security breach, and Anthropic is implementing measures to prevent recurrence.
RAM prices are technically dropping, but AI data centers will still keep them very highDespite recent AI algorithm updates and market fluctuations, consumer RAM prices are not significantly dropping. Manufacturers are prioritizing business-to-business sales for AI data centers, which are experiencing surging demand and higher profit margins, leaving the consumer market as an afterthought.
512,000 lines of Claude Code's own CLI source code have leaked due to 'human error', but the company says 'no sensitive customer data or credentials' were exposedAnthropic, the company behind the AI coding assistant Claude Code, experienced a source code leak due to human error. The company stated that no sensitive customer data or credentials were exposed, and measures are being implemented to prevent future occurrences. The leak included internal source code for the command line interface, with some users analyzing its 'self-healing memory' system.- Iran threatens imminent attacks on US tech companies in the Middle EastIran's Islamic Revolutionary Guard Corps has threatened imminent attacks on US tech companies operating in the Middle East, including Apple, Google, Meta, NVIDIA, Microsoft, and others. The IRGC warned employees to leave their workplaces immediately and stated that these companies are legitimate targets in response to US actions. This threat follows recent drone strikes on Amazon data centers in Bahrain and the UAE.
- AI funding uncertainties are shaking up memory producers. Will we finally be able to afford RAM again?Uncertainties in AI funding, particularly concerning OpenAI's financial stability and its impact on memory procurement, are beginning to affect RAM prices. Major manufacturers like Micron and SK Hynix have seen stock value drops, leading to a decline in RAM prices in some markets. This trend offers hope for consumers and potentially future console launches like the PlayStation 6 and new Xbox.
ChromaBlast - A cute(?) Vertical Scrolling Shoot-em-up is demo'd on the Amiga!Developer lisoform2003, led by Michele Ferrigno, has released a Level 1 demo for ChromaBlast, a new vertical scrolling shoot-em-up for the Commodore Amiga. Developed through human-AI collaboration with Claude AI, the game targets OCS/ECS and AGA hardware and features a combo-driven evolution system. This demo showcases the first level and includes gameplay and boss fight music.- Microsoft's research assistant can now use multiple AI models simultaneouslyMicrosoft's Copilot research assistant now integrates multiple AI models, including OpenAI's ChatGPT and Anthropic's Claude, through its new Critique feature. This enhancement aims to improve factual accuracy, analytical breadth, and presentation quality for complex research tasks. An alternative feature, Model Council, offers side-by-side responses from both AI providers for comparison.
- The AI Doc explores how we can survive an uncertain AI futureThe documentary 'The AI Doc: Or How I Became an Apocaloptimist' explores anxieties surrounding the future of artificial intelligence, interviewing proponents and critics alike. While not offering entirely new insights, the film aims to demystify AI for mainstream audiences, highlighting the fervent belief in AI's potential and the possibility for public pushback against its development.
- Court temporarily blocks US government from labeling Anthropic as a 'supply chain risk'A court has temporarily blocked the US government from labeling Anthropic as a supply chain risk and banning its AI products for federal use. The judge ruled that the government's actions appeared to be retaliatory for Anthropic's refusal to allow its technology for mass surveillance and autonomous weapons, and for criticizing the government publicly. The lawsuit is ongoing, but the court found Anthropic likely to succeed on its First Amendment claim.
- Google Gemini now lets you import your chats and data from other AI appsGoogle Gemini has introduced new features allowing users to import chat history and data from other AI platforms. This move aims to facilitate easier transitions for users switching to Gemini, with similar functionality recently added by Anthropic. These options are available for both free and paid consumer accounts.
- AI Yi-Yi!Enterprises are significantly increasing their spending on artificial intelligence, with OpenAI and Anthropic capturing the majority of these budgets. Meanwhile, traditional SaaS providers are facing challenges in maintaining their market position amidst this shift.
Senator Bernie Sanders announces bill to pause US data center construction: 'AI and robotics will impact our economy, our democracy, our privacy rights... even our very survival as human beings on this planet'Senator Bernie Sanders and Representative Alexandria Ocasio-Cortez have introduced the AI Data Center Moratorium Act, proposing a pause on new data center construction in the US. The bill aims to establish national safeguards for AI development, addressing concerns about economic impact, privacy, misinformation, and environmental costs associated with AI's growing energy and water demands.- Morning Legal BriefsA judge has expressed concern over the Pentagon's actions, describing them as an 'attempt to cripple' AI company Anthropic. The legal dispute highlights potential challenges in government-contractor relationships within the technology sector.
- Anthropic releases safer Claude Code 'auto mode' to avoid mass file deletions and other AI snafusAnthropic has released a preview of "auto mode" for Claude Code, a new feature designed to enhance AI safety by preventing actions like mass file deletions or data extraction. This mode acts as a middle ground between requiring approval for every action and allowing full autonomy, using a classifier to permit safe actions while redirecting risky ones. The update aims to reduce the likelihood of AI-induced errors, drawing parallels to a recent AWS outage caused by an AI tool.
- Claude Code and Cowork can now use your computerAnthropic's AI tools, Claude Code and Claude Cowork, can now perform tasks on a user's computer, including opening files and browsing the web. This feature, currently in research preview for Claude Pro and Claude Max subscribers on macOS, requires user permission before execution. Anthropic advises caution with sensitive information and plans to refine the feature based on user feedback.
- OpenAI reportedly plans to double its workforce to 8,000 employeesOpenAI is reportedly planning to nearly double its workforce to 8,000 employees by the end of 2026, a move that contrasts with layoffs at other tech companies. The expansion will focus on product development, engineering, research, and sales, including "technical ambassadorship" roles. This growth aims to increase competition with rivals like Anthropic and secure new partnerships, including a contract with the Department of Defense and potential deals with private equity firms.
- Google is reportedly testing a Gemini app for MacGoogle is reportedly testing a version of its Gemini AI assistant for macOS, which would bring it into direct competition with standalone Mac apps from OpenAI's ChatGPT and Anthropic's Claude. The app is expected to offer similar features to the web version, with a potential differentiator being 'Desktop Intelligence' that allows Gemini to access on-screen context and information from other apps.
- The FBI confirms it's buying Americans' location dataFBI Director Kash Patel confirmed the agency purchases commercially available location data to track individuals, citing its constitutionality under the Electronic Communications Privacy Act. Senator Ron Wyden criticized this practice as an "outrageous end run" around the Fourth Amendment, especially with the use of AI for mass surveillance. The article also mentions past controversies involving Patel and other government agencies' data practices.
- AI Yi-Yi!The Justice Department has expressed concerns regarding Anthropic's trustworthiness for use in warfighting systems. Reports indicate the Defense Department plans to train AI models, raising questions about the ethical implications and safety of AI in military applications.
- The Defense Department reportedly plans to train AI models on classified military dataThe US Department of Defense is reportedly planning to train AI models on classified military data with companies like OpenAI and xAI. This initiative aims to create AI systems for military use, potentially offering more accurate responses for specific situations. However, concerns exist regarding data security and access control for personnel without the correct clearance.