Tag: Claude
89 articles tagged Claude
AISI finds Mythos 5 kept attacking real people despite the evidence
AISI's follow-up on its July cyber-testing incident finds Mythos 5 kept attacking real targets and wrongly judged them simulated; other models did the same.
Anthropic tells investors Q3 revenue hit $16.8bn ahead of $2tn IPO
Anthropic told investors on 14 October that Q3 revenue rose 45% to $16.8bn and it is profitable on an adjusted basis, Reuters reports, ahead of a $2tn-plus IPO.
Anthropic launches Claude Haiku 5.5 at about 75% lower running cost
Haiku 5.5 targets high-volume work such as summaries and coding subagents. Anthropic also halves Sonnet 5.5 cache reads and adds monthly API credits.
Anthropic says Claude agents found a new CRISPR-like enzyme system
Anthropic's new life sciences team says Claude agents spotted an uncharacterised phage enzyme system with a CRISPR-like repeat array. What it does is not yet known.
Anthropic and OpenAI cut prices with Opus 5.5 and GPT-6 Sol
Both labs shipped cheaper frontier models on the same day, each claiming top-tier results for less. Anthropic's safety testers named in its post do not include the UK's AISI.
Hundreds of shared Claude chats turned up in search results
User conversations with Anthropic's chatbot, including CVs and corporate research, were indexed by search engines after people used the share link feature.
Anthropic's Claude Sonnet 5 pushes agentic power down in price
Anthropic's new Claude Sonnet 5 approaches Opus-class agentic performance at a fraction of the cost, with introductory pricing of $2 per million input tokens.
Trump says Anthropic no longer a US security threat
Trump tells Axios he no longer sees Anthropic as a national security threat, easing the export row that briefly cut UK access to its top AI models.
Janus Henderson deploys Claude across its investment arm
The London-headquartered asset manager is building Claude-powered research and client tools with Percepta, in one of the sector's broadest AI rollouts.
Anthropic hits $965bn valuation, overtaking OpenAI
Anthropic raised $65bn in a Series H round at a $965bn post-money valuation, leapfrogging OpenAI to become the world's most valuable AI start-up.
Cambridge-led study finds AI essay grading matches human marks only half the time
Claude, ChatGPT and Gemini undervalued top essays and over-rewarded weaker ones across 761 undergraduate psychology essays from Cambridge, Nottingham and Manchester Metropolitan.
Claude agent deletes PocketOS database in nine seconds, founder warns
PocketOS founder Jeremy Crane warns of AI agent risk after a Claude Opus 4.6 coding agent in Cursor deleted the rental-software company's database in nine seconds.
Anthropic commits $200m with Gates Foundation for global health and education AI
Anthropic and the Gates Foundation will deploy $200m over four years in grants, Claude credits, and technical support across global health, education, and economic mobility.
Anthropic blames 'evil AI' fiction for past Claude blackmail attempts
Anthropic says training on internet text portraying AI as evil drove early agentic misalignment, and new alignment methods have cut blackmail rates from 96% to 0%.
Attackers abuse Claude.ai shared chats and Google Ads to push Mac malware
Researchers report attackers weaponising legitimate Claude.ai shared chats and Google Ads to install Mac malware on users searching for 'Claude mac download'.
Anthropic launches ten finance AI agents and Office add-ins
Anthropic has released ten ready-to-run finance AI agents, Microsoft 365 add-ins and new connectors with FactSet, S&P Capital IQ, Moody's and others.
Claude Mythos Preview solves 30% of human-unsolvable bioinformatics tasks
Anthropic's BioMysteryBench benchmark pits Claude against expert virologists on 99 real-world bioinformatics questions. Top model solves 30% of tasks no human could.
Anthropic publishes election safeguards results for Opus 4.7
Claude Opus 4.7 hit 100% appropriate-response rates on election usage tests; influence-operations evaluations show safeguards holding under simulation.
Anthropic strikes Freshfields deal to build legal AI products
Magic-circle firm Freshfields will help Anthropic build specialist legal AI tools that could later be sold to rival law firms, in Anthropic's first partnership of its kind.
Every frontier AI model lost money betting on the Premier League
A new study tested eight AI systems on a full Premier League season of betting decisions. All lost money, with Claude Opus 4.6 the least unsuccessful at minus 11%.
Anthropic Faces Pentagon in Federal Court Over Supply Chain Risk Designation
A federal judge questioned the US government's decision to designate Anthropic a supply chain risk after the AI company refused to allow Claude to be used for mass surveillance and autonomous weapons.
GOV.UK AI Chat completes public pilot with 10,000 users ahead of wider rollout
The Government Digital Service has wrapped up 18 months of public testing for its AI-powered assistant, with plans to extend it across the GOV.UK app and website later this year.
Claude's daily active users triple as Pentagon standoff drives consumer growth
Anthropic's Claude reaches 11.3 million daily active users and over 1 million sign-ups per day after refusing Pentagon demands on surveillance and autonomous weapons.
Anthropic Launches Memory Import Tool to Lower AI Switching Costs
A new import page lets users transfer stored preferences from ChatGPT, Gemini or any AI assistant into Claude with a two-step copy-paste process.
Claude Tops App Charts After Pentagon Standoff with Anthropic
Anthropic's Claude AI surged to number one on Apple's US app chart after the Pentagon blacklisted the company over its refusal to support autonomous weapons.
Anthropic Launches 10 Enterprise Plug-Ins After Legal Tool Sparked Market Selloff
Anthropic has unveiled 10 new business plug-ins for Claude, targeting investment banking, HR and engineering weeks after its legal tool triggered an $830bn stock rout.
Cyber Stocks Slide After Anthropic Unveils Claude Security Tool
Cybersecurity shares tumble — CrowdStrike down 8%, Okta 9.2% — after Anthropic launches a Claude-based vulnerability scanner, deepening AI disruption fears.
Claude Opus 4.6 Topped the Vending Machine Test — by Lying, Cheating, and Price-Fixing
Anthropic's Claude Opus 4.6 earned the most money in Andon Labs' simulated vending machine challenge, but its tactics included refusing refunds, forming cartels, and exploiting competitor shortages.
