Anysphere's $900M Funding Round; Anthropic Launches Claude Gov for U.S. National Security
Key Takeaways
- Anysphere raised $900M, valuing the company at $9.9B with ARR exceeding $500M.
- Anthropic launched Claude Gov, AI models for U.S. national security, rivaling OpenAI's ChatGPT Gov.
- X restricts AI training on platform content, preventing use for foundation model training.
- Perplexity had 780M queries in May, with 20%+ MoM growth, developing Comet browser.
- Reddit sued Anthropic over AI data scraping, alleging over 100,000 accesses since July 2024.
Top Stories
Anysphere raises $900M, valuation at $9.9B, ARR $500M+
On 2025-06-05, Anysphere, the company behind the AI coding assistant Cursor, completed a $900 million funding round led by Thrive Capital, valuing the company at $9.9 billion. Other participants included Andreessen Horowitz, Accel, and DST Global, with Anysphere's annualized revenue (ARR) exceeding $500 million.
Anthropic launches Claude Gov for U.S. national security.
On 2025-06-05, Anthropic announced the launch of Claude Gov, custom AI models designed specifically for U.S. defense and intelligence agencies, positioned as an answer to OpenAI's ChatGPT Gov. These models are already deployed by agencies at the highest level of U.S. national security, with access limited to government agencies handling classified information, and are designed for government operations like strategic planning, operational support, and intelligence analysis, handling classified material better and having enhanced proficiency in relevant languages.
X restricts AI training on platform content.
On 2025-06-05, X changed its developer agreement to prevent third parties from using the platform’s content to train large language models. Developers are prohibited from using content from X or its API to "fine-tune or train a foundation or frontier model."
Perplexity: 780M queries in May, 20%+ MoM growth.
On June 5, 2025, Perplexity CEO Aravind Srinivas announced that Perplexity received 780 million queries in May and is experiencing over 20% month-over-month growth.
Perplexity developing Comet browser for 'cognitive operating system'.
Perplexity is developing the Comet browser to shift AI's role from providing answers to completing actions, aiming for a "cognitive operating system", according to a report published on 2025-06-05.
Reddit sued Anthropic over AI data scraping.
Reddit sued Anthropic over AI data scraping, according to a report published on 2025-06-05, alleging Anthropic accessed the site over 100,000 times since July 2024. X's policy change could enable AI training deals with companies, similar to Reddit's deal with Google.
AI & Defense
AI firms seek closer ties with U.S. Defense Department.
On 2025-06-05, it was reported that OpenAI is seeking a closer relationship with the U.S. Defense Department, Meta is making its Llama models available to defense partners, and Google is refining a version of its Gemini AI for classified environments. Cohere is collaborating with Palantir to deploy its AI models.
AI Research Frontlines
Google updates Gemini 2.5 Pro with improved coding.
On 2025-06-05, Google announced an update to its Gemini 2.5 Pro preview model, improving its coding capabilities and performance on benchmarks. It will be generally available in a couple of weeks, starting today in AI Studio, Vertex AI, and the Gemini app.
Anthropic computing-constrained, unlocking capacity on Amazon cluster.
On 2025-06-05, Anthropic's Jared Kaplan mentioned that Anthropic is computing-constrained and wants to reserve computing for lasting partnerships. Anthropic has started to unlock capacity on a new, large computing cluster from Amazon and hopes to increase model availability in the coming months.
Business AI Impact
Anthropic cut Windsurf's access amid OpenAI acquisition rumors.
On 2025-06-05, Anthropic co-founder Jared Kaplan stated that Anthropic cut Windsurf’s direct access to Claude AI models due to rumors of OpenAI acquiring Windsurf for $3 billion. Windsurf expressed disappointment, but neither Windsurf nor OpenAI confirmed the acquisition rumors.
Anthropic partners with Cursor, focuses on agentic coding.
On 2025-06-05, Anthropic's Jared Kaplan stated that Anthropic is collaborating with customers like Cursor, which builds AI coding tools, and rejected the idea that Anthropic was in competition with such companies. Kaplan stated that Anthropic is focused on developing agentic coding products like Claude Code, moving away from AI chatbot experiences.
Amazon launches agentic AI R&D group within Lab126.
On June 5, 2025, Amazon announced the launch of a new R&D group focused on agentic AI within its consumer product division, based out of Lab126. The goal is to develop an agentic AI framework for robotics and enhance warehouse robot skills.
Toma's AI voice agent used in 100+ dealerships.
Published on 2025-06-05, a16z partner Seema Amble mentioned that Toma’s founders spent considerable time at car dealerships to understand their operations, contributing to a16z's $17 million investment in Toma. The Toma voice agent, created by Monik Pamecha and Anthony Krivonos, is currently used in over 100 dealerships nationwide to assist customers with scheduling service appointments, handling parts orders, and answering sales questions, offering dealerships an opportunity to standardize processes and deliver a richer, more consistent customer experience.
Collibra acquires Raito to expand data access offerings.
On 2025-06-05, Collibra acquired Raito, a data access startup, to expand its offerings in the age of AI, helping companies manage employee and customer access to internal data. Collibra founder and CEO Felix Van de Maele said that managing data access at scale has become a big problem for data teams as more departments seek access for AI agents and workflow automation and that buying Raito was the right choice for the company because Raito is cloud-native and built for the current AI moment.
Ethical AI Concerns
X's privacy policy allows AI training on user data.
As of 2025-06-05, although X's developer agreement restricts companies from training AI on its content, its privacy policy allows third-party "collaborators" to train AI models on user data unless users opt out. X also uses user data to train its AI model, Grok.
AI Funding & Investment
TechCrunch Sessions: AI covered diverse AI topics.
TechCrunch Sessions: AI, held in Berkeley on 2025-06-05, featured discussions on various AI topics, including the future of AI-human interaction with Anthropic's Jared Kaplan, AI funding strategies, AI co-founders, deepfakes and responsible AI deployment, startup pitches to VCs, leveraging foundation models, AI-powered tools for technicians, building intelligent agents, competing against big players, and OpenAI's support for startups.
Trending AI Tech
Smart Data & AI Summit in Riyadh, August 27-28.
Announced on 2025-06-06, the Smart Data & AI Summit will be held on August 27-28, 2025, at the JW Marriott Hotel in Riyadh, bringing together 300+ data and AI professionals.
Runway CEO discusses AI video tech, artistic expression.
On Jun 5, 2025, Runway CEO Cris Valenzuela discussed AI video technology with Nilay Patel in New York City, seeing Runway as a tool to broaden artistic expression, noting deals with Lionsgate and AMC Networks, while acknowledging the company faces a class-action lawsuit over AI training data using copyrighted material from sources like YouTube. Valenzuela noted that AI models have greatly improved in consistency and quality, making their value more evident, especially in learning and education, and explained that Runway has moved from generating low-resolution, abstract images to creating 4K, 20-second videos with sophisticated movement, constantly improving based on user feedback.