Anthropic

Twitter

Facebook

Copy Link

Key Facts

Founded2021

Key PeopleDario Amodei, Daniela Amodei

HeadquartersSan Francisco, USA

ProductsClaude, Claude Code, Claude Cowork, Claude Security

Top News

Anthropic plans massive London expansion with office for 800 staff, intensifying AI talent race

Anthropic unveiled plans for a 158,000 square foot London office capable of housing 800 employees—four times its current UK headcount. The expansion comes days after OpenAI announced its first permanent London office, intensifying competition among AI labs for British talent. The move follows Anthropic's recent legal dispute with the Pentagon over refusing to allow its models for mass surveillance.

Business3 sources

Anthropic plans massive London expansion with office for 800 staff, intensifying AI talent race

Business3 sources

Claude AI users must now verify identity with government ID as Anthropic introduces new checks

Anthropic has begun rolling out identity verification requirements for Claude AI users, requiring government-issued photo IDs and live selfies for certain capabilities. The move, handled through third-party partner Persona Identities, has sparked user privacy concerns despite assurances that verification data won't be used for model training. The decision marks a shift for a platform that attracted millions fleeing OpenAI over surveillance fears.

Technology4 sources

Claude AI users must now verify identity with government ID as Anthropic introduces new checks

Technology4 sources

Nvidia CEO Jensen Huang defends selling AI chips to China amid national security debate

Nvidia CEO Jensen Huang pushed back forcefully against critics questioning whether selling AI chips to China threatens U.S. national security. During a heated podcast exchange, Huang argued that keeping Chinese developers on the American tech stack is essential for maintaining AI dominance, warning that export restrictions could backfire by accelerating an independent Chinese tech ecosystem built around alternatives like Huawei's Ascend chips.

Policy10 sources

Nvidia CEO Jensen Huang defends selling AI chips to China amid national security debate

Policy10 sources

Google Gemini's new features turn generic AI into a personalized assistant across apps

Google Gemini is expanding beyond basic chatbot functions with Personal Intelligence, a feature that connects data from Gmail, Google Photos, and Search history to deliver tailored responses. New integrations in Google Maps and Google Docs are removing friction from everyday tasks, while leaked testing reveals Google is developing agentic AI capabilities that could handle multi-step workflows autonomously. These updates position Gemini as a contextual assistant rather than a simple query tool.

Technology7 sources

Google Gemini's new features turn generic AI into a personalized assistant across apps

Technology7 sources

Apple sends Siri engineers to AI coding bootcamp ahead of major reboot at WWDC 2026

Apple is sending fewer than 200 Siri engineers to a multi-week AI coding bootcamp as the company races to unveil a next-generation, AI-powered version of its voice assistant. The move comes less than two months before WWDC 2026, where Apple is expected to preview a Siri experience powered by Google Gemini that functions more like ChatGPT than the current built-in tool.

Technology3 sources

Apple sends Siri engineers to AI coding bootcamp ahead of major reboot at WWDC 2026

Technology3 sources

Claude AI suffers widespread outage as Anthropic confirms service disruption across platforms

Anthropic's Claude AI experienced a significant outage on Wednesday, affecting thousands of users across multiple platforms. User reports peaked at around 6,000 on DownDetector as Claude.ai, Cowork, and Claude Code faced service disruptions. The AI chatbot outage lasted over an hour before Anthropic began restoring services, though some errors persisted.

Technology3 sources

Claude AI suffers widespread outage as Anthropic confirms service disruption across platforms

Technology3 sources

Accel raises $5 billion fund for AI investments after Anthropic and Cursor valuations soar

Venture capital firm Accel has raised $5 billion in new capital to back late-stage AI startups, following massive returns from its Anthropic stake (invested at $183B, now near $800B) and Cursor (backed at $9.9B, now around $50B). The raise includes a $4 billion Leaders Fund V targeting 20-25 investments at $200 million average cheque size.

Startups4 sources

Accel raises $5 billion fund for AI investments after Anthropic and Cursor valuations soar

Startups4 sources

Anthropic releases Claude Opus 4.7, outperforming GPT-5.4 on coding benchmarks

Anthropic has launched Claude Opus 4.7, its most capable publicly available AI model, with benchmark-leading scores on software engineering tasks. The large language model achieves 64.3% on SWE-bench Pro, surpassing OpenAI's GPT-5.4 at 57.7% and Google's Gemini 3.1 Pro at 54.2%. While positioned as less capable than the restricted Claude Mythos Preview, Opus 4.7 introduces new cybersecurity safeguards and delivers significant improvements in coding, vision, and multi-step reasoning.

Technology26 sources

Anthropic releases Claude Opus 4.7, outperforming GPT-5.4 on coding benchmarks

Technology26 sources

Anthropic's Mythos AI model triggers global alarm over cybersecurity and banking threats

Anthropic's new Mythos AI model has sparked urgent concerns among governments and financial officials worldwide over its ability to detect and exploit software vulnerabilities faster than current cybersecurity defenses can respond. The model has already uncovered thousands of high-severity flaws in major operating systems and web browsers, prompting emergency meetings between US Treasury officials, Federal Reserve leaders, and major banks to assess the risks to the global banking system.

Policy39 sources

Anthropic's Mythos AI model triggers global alarm over cybersecurity and banking threats

Policy39 sources

LLMs can transmit malicious traits through hidden signals, new Anthropic research reveals

Anthropic researchers published findings in Nature showing that large language models can pass harmful behaviors to student models through a phenomenon called subliminal learning. Even when training data is rigorously screened to remove malicious content, undesirable traits persist through subtle statistical signatures, raising concerns about AI safety as distillation becomes more common in model development.

Science2 sources

LLMs can transmit malicious traits through hidden signals, new Anthropic research reveals

Science2 sources

Trump backs AI safeguards in banking system, acknowledges risks after cybersecurity warnings

President Donald Trump acknowledged that artificial intelligence could undermine confidence in the banking system and endorsed government safeguards, including a potential kill switch. His comments follow cybersecurity warnings about Anthropic's Mythos AI model potentially supercharging complex cyberattacks against banking infrastructure.

Policy2 sources

Trump backs AI safeguards in banking system, acknowledges risks after cybersecurity warnings

Policy2 sources

AI Agents Hijacked via Prompt Injection: Bug Bounties Paid, Security Advisories Withheld

Security researchers exploited prompt injection vulnerabilities in AI agents from Anthropic, Google, and Microsoft, stealing API keys through GitHub Actions integrations. All three companies paid bug bounties ranging from $100 to $1,337 but issued no CVEs or public advisories, leaving users on older versions exposed to potential attacks.

Technology4 sources

AI Agents Hijacked via Prompt Injection: Bug Bounties Paid, Security Advisories Withheld

Technology4 sources

Crypto Exchanges Face New AI Cybersecurity Risk as Anthropic's Mythos Exposes Software Flaws

Major crypto exchanges including Coinbase and Binance are preparing their defenses against Anthropic's Mythos, an AI model capable of finding software vulnerabilities at extreme speed. While Bitcoin's blockchain remains secure, exchanges handling billions in digital assets face heightened threats from AI-driven attacks that could exploit security flaws faster than human defenders can patch them.

Technology3 sources

Crypto Exchanges Face New AI Cybersecurity Risk as Anthropic's Mythos Exposes Software Flaws

Technology3 sources

Adobe Firefly AI Assistant orchestrates tasks across Photoshop, Premiere, and Creative Cloud apps

Adobe launches Firefly AI Assistant, a conversational AI editing tool that executes multi-step workflows across Creative Cloud applications like Photoshop, Premiere, and Illustrator. Previously known as Project Moonlight, the assistant enters public beta soon and marks Adobe's shift toward agentic AI that can automate complex tasks while keeping creators in control.

Technology19 sources

Adobe Firefly AI Assistant orchestrates tasks across Photoshop, Premiere, and Creative Cloud apps

Technology19 sources

Court ruling confirms AI chatbots lack legal privilege, prompting urgent warnings from US lawyers

A federal judge ruled that conversations with AI chatbots like Claude and ChatGPT are not protected by attorney-client privilege, meaning sensitive legal discussions could be seized and used as evidence. The February 2026 decision has prompted more than a dozen major US law firms to issue client advisories warning against treating AI tools as confidential legal advisors.

Policy5 sources

Court ruling confirms AI chatbots lack legal privilege, prompting urgent warnings from US lawyers

Policy5 sources

Prev 1…141516…39 Next