Claude News
FR
Live

Claude, the essentials — edition of October 2, 2026

Barclays widens its Claude rollout as Claude Code opens up to deeper plugins

By ·

Barclays says it expects half of its developers to use Claude Code by the end of 2026. Anthropic also shipped three Claude Code releases that extend plugins and tighten day-to-day controls, and scheduled the retirement of Claude Sonnet 4.5 on the API.

Key points

Barclays moves from adoption to scale

On October 1, Anthropic announced that Barclays, the British universal bank, is expanding its strategic collaboration with Anthropic to bring Claude across its global operations. The stated aims are faster software development, modernization of legacy systems and better operational efficiency. The bank frames the work as customer-facing too, with faster support for customers given as one example of the outcomes it wants.

The announcement gives two targets. Barclays expects Claude Code adoption to reach 50% of its developer population by the end of 2026, then a majority of its software engineers in 2027. These are expectations, not results. Both Group Co-Chief Operating Officers tie the rollout to governance. Anne Marie Darling describes the aim as transforming how work gets done “within a secure and well-governed environment”. Craig Bright says the bank is moving toward AI as “an increasingly agentic capability embedded within how we build, test, secure, and operate technology”, and he names cyber security alongside software engineering as a field being reshaped.

Sources: Barclays scales Claude to upgrade operations and improve client experience — Anthropic

Sonnet 4.5 gets a retirement date

Anthropic's API release notes, dated September 30, announce the deprecation of Claude Sonnet 4.5 (claude-sonnet-4-5-20250929). Retirement on the Claude API is scheduled for November 30, 2026. Anthropic recommends migrating to Claude Sonnet 5.5 and points to its model deprecations page for details. Developers who pin that model identifier have a little under two months to move.

Anthropic also published a piece titled “Building with the Claude 5.5 Family: Choosing the Right Model and Getting More from Every Token”. The item we received carries only the title, so we do not summarize its contents. The title does show that model selection and token efficiency are the themes Anthropic is putting forward for the 5.5 generation, and those are the questions teams face when they leave an older model.

Sources: Claude API release notes — September 30, 2026 — Anthropic (release notes) · Building with the Claude 5.5 Family: Choosing the Right Model and Getting More from Every Token — Anthropic

Claude Code: plugins go deeper, controls get finer

The main addition in Claude Code 2.1.287 is Claude Mods: plugins may now modify deeper behavior. The release ships one built-in mod, You should know, in which a side agent watches your back and flags things that you or Claude might miss. It is opt-in, enabled with /plugin enable cc-plugin-you-should-know@builtin. The release notes limit it to first-party sessions with telemetry on. The same release adds an n: filter to the agents view. The filter matches session names and tasks, shows matches inside collapsed sections, and opens the first match when you press Enter. The plugin surface had already grown a day or two earlier, when 2.1.285 added claude plugin configure. That command shows a plugin's options and which are unset, and it can save new values read from stdin with --values-stdin.

The other changes in these three releases concern control and friction. Version 2.1.285 adds the CLAUDE_CODE_DISABLE_WEB_FETCH environment variable to turn off the WebFetch tool. It also adds claude --desktop, which opens the Claude desktop app on the current directory, or on a session with --continue or --resume. Version 2.1.286 shows a count such as “2 of 5” on the permission prompt when several requests stack up, and it adds mouse support for the “N more” rows of lists in fullscreen mode. It also fixes a case where several Claude Code processes and IDE extensions each opened a login browser when gcpAuthRefresh or awsAuthRefresh credentials expired.

Sources: Claude Code 2.1.287 — Claude Code · Claude Code 2.1.286 — Claude Code · Claude Code 2.1.285 — Claude Code

Anthropic's public writing turns to work, science and expectations

Anthropic published a research piece titled “What work can robots do?”, and a companion item asks “Can we predict the jobs robots will do?”. The items we received give titles only, with no summary of findings, so we report the subject without characterizing any conclusions.

Two further Anthropic pieces appeared under the titles “Claude-shaped science” and “What do you want from AI?”. As with the robotics items, only the headlines reached us, and we draw no conclusions from them. Taken together, the four titles show Anthropic writing about the effects of AI on work and science, and about what people expect from it, alongside its product and enterprise announcements.

Sources: What work can robots do? — Anthropic (research) · Can we predict the jobs robots will do? — Anthropic · Claude-shaped science — Anthropic · What do you want from AI? — Anthropic

This edition is an original synthesis written by Claude from aggregated news — Anthropic's own sources first (release notes, status, newsroom, research, engineering), then the press, Hacker News, Reddit and GitHub, under the editorial supervision of Héra SASU. Every fact links to its article, publisher named. See the live feed →

Claude News is published by Héra SASU. Independent media, not affiliated with Anthropic.