@Claude Graduates to a Named Product in Anthropic’s Lineup
The async Slack helper launched June 23 as Claude Tag now has its own product page and a permanent spot in Anthropic’s product navigation as @Claude — sitting alongside Claude Code, Cowork, and Claude Design rather than buried as a feature. You tag @Claude in a Slack thread, hand off a task, and it works asynchronously, even scheduling follow-up work for itself over hours or days.
The positioning matters more than the mechanics. Anthropic is signaling that “Claude where your team already talks” is a first-class surface, not an integration afterthought. For Team and Enterprise customers, it’s another nudge toward treating Claude as a coworker you delegate to, not a chat window you visit.
Monday Status Check: Fable 5 Still Offline, Opus 4.8 Carries the Frontier
The weekend brought no change at the top of Anthropic’s newsroom: the June 12 statement on the US export-control directive suspending Fable 5 and Mythos 5 is still the lead item, with no public clearance for the consumer model. For everyday users and most builders, Opus 4.8 remains the most capable model you can actually call.
It’s a strange holding pattern: the frontier exists but sits behind a government gate, while the workhorse model keeps shipping the real work. If you’re planning a roadmap around frontier capability, build for Opus 4.8 today and treat any Fable return as upside, not a dependency.
Claude Code Ships Artifacts: Sessions Become Live, Shareable Pages
Claude Code can now capture a session’s work as an artifact — a live web page built from your full context: the codebase, your connectors, and the conversation itself. Think PR walkthroughs, incident timelines with error-rate charts, dashboards you can filter, or a release checklist that fills itself in as work completes.
The clever part is that pages update in place: every publish is a new version at the same link, with full version history and a gallery to manage them. Artifacts are private to your org by default and can’t be made public, with admin role-based scoping and retention controls. It’s in beta for Team and Enterprise, from the CLI and desktop app, viewable in any browser.
The Latest Claude Code Drop Is All Polish: Voice, Mouse, Background Agents
No headline feature this time — just a stack of quality fixes. The newest Claude Code notes add a fullscreen mouse-click toggle (CLAUDE_CODE_DISABLE_MOUSE_CLICKS keeps wheel scroll while disabling click/drag/hover), fix voice dictation capturing silence after an input-device switch, and fix auto-submit never firing for languages written without spaces — Japanese, Chinese, Thai.
There’s also steadier background-agent behavior (the launch result no longer tells Claude to end its response, so it keeps working), a remote-session startup checklist while the container provisions, better Linux mic detection, and a hook-matcher fix so hyphenated identifiers exact-match instead of substring-matching. Unglamorous, but exactly the reliability work that makes a daily-driver tool stick.
Uber Torched Its 2026 AI Budget in Four Months — and Every CTO Noticed
Uber burned through its entire 2026 AI coding budget in four months, much of it on Claude Code and Cursor. Adoption ran ahead of every forecast: agentic Claude Code users jumped from 32% of engineers in February to 84% by March, with monthly per-engineer API costs landing between $500 and $2,000. Uber has since capped spend at $1,500 per tool per engineer per month.
The quote that’s traveling: COO Andrew Macdonald saying it’s hard to connect rising Claude Code use to consumer-facing innovation — “that link is not there yet.” This is the story turning into a playbook. When agents bill per token, adoption and cost rise together, and finance notices before the ROI does. Set per-key spend limits and per-seat caps before your own leaderboard does the budgeting for you.
Anthropic Leads the Enterprise — Pricing Is the Soft Spot
The scoreboard reads well for Anthropic. By April, more US businesses paid for Claude than for ChatGPT (34.4% vs 32.3%), Claude Code holds roughly 54% of the enterprise coding market, run-rate revenue has cleared $30 billion, and more than a thousand customers spend over $1M a year. The framing that’s stuck: OpenAI is a consumer company selling enterprise products; Anthropic is an enterprise company that happens to have a consumer one.
But today’s Uber story points at the soft spot. The same usage-based pricing that lets Claude Code go as deep as a problem needs also turns adoption into an uncapped meter — and budget shocks travel fast among CTOs who all talk to each other. Anthropic’s next enterprise battle may be less about capability than about predictability: spend controls, caps, and pricing that a finance team can model a quarter ahead. Win that, and the lead compounds. Ignore it, and “too expensive to standardize on” becomes the competitor’s easiest pitch.