Tag: anthropic
All the articles with the tag "anthropic".
-
Anthropic Passed OpenAI in Q2 Revenue. The Profit Line Is the Real News.
Anthropic's Q2 revenue hit $11.5B, passing OpenAI's $6.7B for the first time. But the first positive adjusted operating income at a frontier lab is the bigger story.
-
Claude for Teachers Is Not a Homework Story
Anthropic's teacher-focused Claude launch points to a better classroom AI frame: less student shortcut, more instructional infrastructure.
-
Microsoft Using Its Own Models Is the AI Cost Reality Check
Microsoft reportedly shifting more Office prompts to in-house MAI models shows the AI race moving from best model to best cost-control stack.
-
Anthropic Talking to Samsung Is the Frontier AI Supply Chain Story
Anthropic reportedly discussing a custom AI chip with Samsung shows frontier labs moving from model access into supply chain control.
-
Fable 5 Came Back, But the Precedent Stayed
Anthropic restored Fable 5 after US export controls were lifted, but the bigger lesson is that frontier model access now has a review loop.
-
Claude Fable 5 Shows The Frontier Model Split Getting Real
Anthropic's Fable 5 and Mythos 5 launch shows a practical frontier AI pattern: broad access for guarded capability, restricted access for higher-risk power.
-
Biology Agents Need Deterministic Tools, Not Just Smarter Models
Anthropic's biology-agent work shows why scientific AI needs reliable data retrieval layers before agents can safely operate in messy research infrastructure.
-
Claude Code Dynamic Workflows Turn Agents Into Swarms
Anthropic's Opus 4.8 and dynamic workflows show where coding agents are going: not one assistant, but coordinated agent swarms with verification.
-
Anthropic Just Said the Quiet Part About Agent Security
Anthropic's Claude containment post makes the real agent-security lesson obvious: permission prompts are not enough. The boundary has to be deterministic.
-
Project Glasswing Moved the Bottleneck From Finding Bugs to Fixing Them
Anthropic says Claude Mythos Preview and roughly 50 partners found more than ten thousand high- or critical-severity vulnerabilities. The scary part is not discovery anymore. It is disclosure, triage, and patch throughput.