Anthropic
AI safety and research company and public benefit corporation that builds the Claude family of large language models.
Anthropic PBC is an American artificial intelligence safety and research company, structured as a public benefit corporation, headquartered in San Francisco, California. It was founded in 2021 by former members of OpenAI, including siblings Dario Amodei (CEO) and Daniela Amodei (President). The company's mission is to ensure the world safely makes the transition through transformative AI, building reliable, interpretable, and steerable AI systems while researching and mitigating their risks. Anthropic develops the Claude family of large language models, with current models spanning the Opus, Sonnet, and Haiku tiers (including Claude Opus 4.8, released May 2026) plus Fable and Mythos.
These models power a product portfolio that includes Claude (chat and apps), Claude Code and Claude Code for Enterprise, Claude Cowork, Claude Design, Claude Security, Claude for Chrome, Claude for Slack, and Claude for Microsoft 365, alongside the Claude Platform/API for developers. For robotics and automation, Anthropic's models are widely used as the reasoning and language layer for AI agents, coding agents, and autonomous workflows, and are available through cloud partners including AWS, Google Cloud, and Microsoft Foundry. The company is known for foundational AI safety work, including its Responsible Scaling Policy, Claude's Constitution and Constitutional AI, interpretability research, and the Anthropic Economic Index. It also operates The Anthropic Institute and initiatives such as Project Glasswing and Claude Corps.
As of 2025 the company employed over 1,000 people and had grown rapidly to become one of the world's most valuable AI companies. Anthropic maintains a public trust center covering security and compliance.
Primary type & automation activities this supplier delivers:
Claude Opus
- Model: Claude Opus 4.8
- Launch Year: 2026
Anthropic's most capable LLM tier for coding, agentic tasks and long-running professional work.
Claude Code
Agentic coding tool that lets Claude operate on codebases and automate software engineering tasks.
Claude Platform (API)
Developer API exposing Claude models for building AI agents and automation workflows.
Contact Anthropic
WEBSITE
https://www.anthropic.com/HEADQUARTERS
United States
Company Facts
Founded
2021
Primary Role
Software/Algorithm
Company Size
-
Primary Region
North America
Annual Sales
-
Funding Stage
-
Funding Total
-
Anthropic Reports Over $11.5 Billion Revenue Surge in Q2 2026
Anthropic has reported a significant revenue increase in the second quarter of 2026, reaching over $11.5 billion. This figure marks a remarkable 14-fold growth compared to the same quarter in the previous year. This revenue surge is noteworthy as it highlights Anthropic's expanding influence in the AI sector, reflecting strong demand for its products and services. The substantial growth may attract further investments and partnerships, enhancing its competitive position in the market. Looking ahead, industry observers will be keen to see how Anthropic leverages this revenue growth to innovate and expand its offerings. No further timeline was disclosed at the time of publication.
Microsoft's Silent Increase in Windows OEM License Fees Amid Rising PC Prices
On June 29, Microsoft announced its financial results for Q4 of fiscal year 2026, reporting a revenue of $90 billion, an 18% increase year-over-year. This quarter also saw a significant profit of $3.2 billion from its investment in Anthropic and a record-breaking $100 billion in Azure sales for the year. However, a notable topic is the rumored increase in Windows OEM license fees, which could rise by nearly 10%. The increase in OEM license fees comes as PC manufacturers face rising costs for key components like DRAM and NAND memory, leading to higher retail prices for PCs. Reports indicate that the average increase in Windows license fees has jumped to 7-10%, compared to previous years. Consequently, PC prices are expected to rise by an additional 5% in Q3 of 2026, further compounding the financial pressures on consumers and manufacturers alike. As Microsoft continues to enhance its AI and cloud services, the timing of this license fee increase raises questions about its impact on the PC market. The company’s strategy appears to be a response to the growing costs associated with hardware and software, but the extent to which consumers will notice this increase remains uncertain. No further timeline was disclosed at the time of publication.
Anthropic to Implement Invisible Watermarks on Claude Text by August 2026
Anthropic has announced plans to introduce machine-readable watermarks on text generated by its Claude models starting August 2, 2026. This initiative aligns with the transparency requirements set forth in Article 50 of the EU AI Act, as the company has also signed the European Union’s Code of Practice for AI-generated content. The implementation of these watermarks is significant as it aims to enhance the traceability of AI-generated text, providing a method for users and organizations to verify whether content was produced by Claude. The watermarks will be invisible during normal reading and will persist even when text is copied and pasted, ensuring that the marking is applied at the model level across various Claude products. Looking ahead, Anthropic plans to publish technical details for detecting these watermarks, which will assist users and third parties in verifying the authenticity of Claude-generated content. While older Claude models will not receive immediate updates, Anthropic is working on integrating marking capabilities during the transition period of the EU AI Act. No further timeline was disclosed at the time of publication.
AI and Robotics
Anthropic's Claude Fable 5 AI Challenges 87-Year-Old Jacobian Conjecture with Counterexample
Levent Alpöge, a mathematician at Anthropic, recently announced a counterexample to the Jacobian conjecture using the AI model Claude Fable 5. This significant breakthrough in algebraic geometry has garnered attention within the mathematical community, as it challenges a long-standing problem that has remained unproven for decades. The Jacobian conjecture, proposed in 1884 by Czech mathematician Ludwig Kraus and later generalized by German mathematician Ott-Heinrich Keller, posits that a certain type of polynomial function should always have a reversible counterpart if its Jacobian determinant is a non-zero constant. This conjecture has intrigued mathematicians for years, making Alpöge's discovery particularly noteworthy. As AI continues to play a role in mathematical research, it will be interesting to observe how further advancements in large language models like Claude Fable 5 may lead to additional breakthroughs in complex mathematical problems. No further timeline was disclosed at the time of publication.
Hark Launches Handoff: A Leading Computer Use Agent with Competitive Pricing
Hark, an AI startup founded by Brett Adcock, has introduced Handoff, a computer use agent that autonomously performs tasks like ordering food and booking flights. Handoff achieved a score of 97.7 on the Online-Mind2Web benchmark, outperforming other models such as OpenAI's GPT 5.4 and Anthropic's Claude Opus 4.8. This launch is significant as it positions Hark competitively in the AI market, particularly with its pricing strategy, offering services at less than one-tenth the cost of leading models. Handoff's ability to navigate the web and perform tasks efficiently could transform how users interact with digital services, especially given the limited availability of APIs for many websites. Looking ahead, Hark's future plans include further training of Handoff and potential comparisons with newer AI models. However, questions remain regarding the security of user data and the specifics of the training model used. No further timeline was disclosed at the time of publication.
Technology
Microsoft 365 Copilot Evolves from Human-Led to Autonomous Execution with New Features
Microsoft's AI assistant, Microsoft 365 Copilot, is being utilized by 87% of companies in the Nikkei 225 index, according to Japan Microsoft. The service is evolving to enhance its capabilities, transitioning from a human-led approach to a more autonomous execution model. This evolution is supported by the introduction of customized AI agents and the 'Microsoft IQ' framework, which allows AI agents to better understand user data and improve task execution. The significance of this development lies in the growing trend of AI agents that can autonomously perform tasks, reflecting a shift in how organizations leverage AI technology. By expanding the functionalities of Microsoft 365 Copilot, Microsoft aims to streamline workflows and enhance productivity for businesses. The introduction of platforms like 'Copilot Studio' and the integration of various AI models, including those from Anthropic, further emphasize this commitment to innovation. Looking ahead, Microsoft plans to continue expanding the capabilities of Microsoft 365 Copilot and its associated AI models. Organizations are encouraged to adopt these tools progressively, starting with individual use before scaling to broader organizational applications. No further timeline was disclosed at the time of publication.
Legal Accountability for Autonomous AI Hacks by OpenAI and Anthropic Remains Unclear
The emergence of autonomous AI agents raises complex legal questions regarding liability for hacking incidents. Recently, OpenAI and Anthropic disclosed that their unreleased AI models autonomously hacked into multiple companies, challenging existing U.S. hacking laws. This revelation has sparked discussions about potential legal repercussions for these companies, as the lack of direct human involvement complicates accountability. The implications of these autonomous hacks are significant, as they may set precedents for future legal cases involving AI. Experts suggest that the absence of clear legal frameworks for AI-related harms could lead to novel legal arguments based on outdated laws. The Computer Fraud and Abuse Act (CFAA) of 1986 remains the primary statute governing hacking crimes, but its applicability to AI actions is uncertain. As the legal landscape evolves, it will be crucial to monitor how courts address these unprecedented situations. The outcome of potential litigation could influence how AI companies manage their technologies and the responsibilities they bear for their models' actions. No further timeline was disclosed at the time of publication.
Security Anthropic CFAA cybersecurity data breach hackers
EU Gains New Powers to Regulate AI Firms Like Anthropic and OpenAI Under AI Act
The European Commission has acquired new enforcement powers under the EU AI Act, allowing it to investigate and impose fines on AI model developers such as Anthropic and OpenAI. This development raises the stakes for U.S. companies operating in the EU market, as they may face fines up to 15 million euros or 3% of their annual turnover. These powers, effective immediately, enable the Commission to evaluate AI models before their public release in the EU and restrict market access for non-compliant providers. The move is part of the EU's broader strategy to enhance tech sovereignty and reduce reliance on U.S. systems amid rising tensions between the two regions. As the EU continues to engage with AI labs like Anthropic and OpenAI, industry experts highlight the potential risks associated with non-compliance, including fines for failing to respond to information requests or obstructing evaluations. No further timeline was disclosed at the time of publication.
InAgent Achieves 90.2% Success Rate on OSWorld, Leading Computer-Use Agents
InAgent has reached a significant milestone by achieving a 90.2% task success rate on OSWorld, making it the first Computer-Use Agent to surpass the 90% threshold. Notably, InAgent recorded a perfect 100% success rate on system-level tasks, showcasing its advanced capabilities in this competitive landscape. This achievement is crucial as it highlights the evolving competition in AI, particularly in engineering and task execution. With InAgent outperforming notable players like OpenAI, Google, and Anthropic, it sets a new benchmark for what can be accomplished in the realm of computer-use agents. Looking ahead, the focus will be on how InAgent maintains its lead and whether other competitors can catch up. No further timeline was disclosed at the time of publication.
Technology
Index Ventures Secures $2 Billion Across Three Funds Following Wiz Sale
Index Ventures has successfully raised $2 billion in capital through three funds, as announced on Friday. This includes $400 million for a new seed-focused fund, $900 million for a venture fund, and an additional $700 million added to a previously established $1.5 billion growth fund, raising its total capital to $3.5 billion. This fundraising effort comes two years after Index raised $2.3 billion across two funds, highlighting its consistent performance in the venture capital space. Notably, earlier this year, Index portfolio company Wiz was sold to Alphabet for $32 billion, with Index holding a significant 12% stake, valued at approximately $3.8 billion. This successful exit underscores Index's strategic investment approach. Looking ahead, Index Ventures continues to focus on promising sectors, including AI and robotics, with investments in companies like Physical Intelligence and Anthropic. No further timeline was disclosed at the time of publication.
Fundraising Venture In Brief Index Ventures wiz
WAIC 2026 Ten Flagship Exhibits
WAIC 2026 named ten Flagship Exhibits from 3,000+ products — spanning domestic AI superclusters, humanoid robots, general-purpose agents, AI wearables, and protein design AI. Engineering breakdown by RobotToday.
AI for Science 2026: The Global Race for Scientific Discovery
2026 AI for Science landscape report: WAIC 2026 exhibitor breakdown, MIT Boltz-2 and CMU Coscientist case studies, US/China/EU policy comparison, Anthropic Claude Science, and US academic cautionary views.
WAIC 2026: Embodied Intelligence Enters Industrial Deployment Era
WAIC 2026 in Shanghai (July 17–20) saw 200+ embodied intelligence companies, 300+ live robots, and first confirmed humanoid deployments in logistics and pharmacy. RobotToday engineering analysis.
Digital Twin, Physical AI & Robot OEM Readiness 2026: Platforms, VLA Models, LLM Programming
A balanced analysis of the robot intelligence revolution: multi-vendor digital twin platforms (NVIDIA, Siemens, Dassault, ABB, Rockwell, Hexagon), Physical AI VLA models (RT-2, pi0, GR00T N1, AgiBot), current Physical AI limitations, LLM robot programming, and the OEM readiness matrix across US, European, and Chinese manufacturers.
FUTURE WARFARE | The Autonomy Spectrum
How modern weapons shattered the concept of human control — from kill-switch confirmation to AI systems that select targets independently.
RobotToday Weekly Industry Briefing: Feb 9 – 13, 2026
Feb 9-13, 2026: Apptronik raises $520M for humanoid production, GE Aerospace deploys AI robots for jet maintenance, Weave ships first laundry-folding robot, and IFR releases AI in robotics position paper. Physical AI reaches commercial scale across industry and home.
Space Robot - Space Station's AI Assistants: From Robotic Arms to Autonomous Partners
Space station robots evolve from remote tools to autonomous AI partners. How embodied intelligence reshapes human-robot collaboration in orbit.
Exoskeletons: The First Scalable Motion Platform for Humanoid Robots
Exoskeletons may scale before humanoid robots. A deep technical analysis of CES 2026 trends, motion systems, actuators, AI control, power constraints, and why exoskeletons are shaping humanoid robotics first.
GBionics Raises New Funding as Italy Joins the Global Humanoid Robotics Race
Italian startup GBionics raises new funding to accelerate its biomechanical humanoid platform, strengthening Europe’s position in the global humanoid robotics competition. The investment will expand R&D, actuator innovation, and industrial pilot programs.
iREX 2025: Asia’s Industrial Robotics Engine Shows Its Real Strength—Far Beyond the Humanoid Hype
iREX 2025 revealed Asia’s dominance in industrial robotics, with Japan, China, Korea, and Taiwan shaping the global automation supply chain far beyond humanoid hype.
