Anthropic commits $100M to the engineers who deploy Claude as Barclays scales up

Enterprise AI meets a talent question

On October 2, Anthropic launched Claude Frontier Academy, backed by $100 million, with the aim of training 10,000 Frontier Deployed Engineers by the end of 2027. Anthropic locates the bottleneck in people, not technology. It says every enterprise is racing to bring AI in, that the people able to make it work inside a real business have become the hardest talent to find, and that a small group of deeply skilled people drives an outsized share of what AI delivers. The first program, the Frontier Deployed Engineer Residency, follows what Anthropic calls the medical model. Doctors learn from practicing physicians, train on real cases and are assessed before they practice alone, and residents are meant to learn from practitioners in the same way. Anthropic describes the Academy as the first of its kind from an AI company, built from its own work deploying Claude inside large enterprises and the professional-services firms that serve them. Steve Corfield, Anthropic's Global Head of Business Development and Partnerships, says the Academy “trains people the way our own engineers learn, and we want those who graduate to set the standard for how AI gets built inside a business.”

The day before, on October 1, Barclays announced an expanded collaboration with Anthropic. The bank is extending Claude across its operations to speed up software development, modernize legacy systems and improve operational efficiency. It expects Claude Code adoption to reach 50% of its developer population by the end of 2026 and a majority of its software engineers in 2027. Craig Bright, Barclays' Group Co-Chief Operating Officer, describes a move toward AI as an increasingly agentic capability embedded in how the bank builds, tests, secures and operates technology. His co-COO Anne Marie Darling stresses a secure and well-governed environment. The two announcements are separate, and the material does not link Barclays to the Academy. Read side by side, they show the same dynamic from two angles: large organizations are committing to broad deployment, and Anthropic is investing in the skills needed to carry it out.

Sources: Claude Frontier Academy: $100M to train 10,000 engineers — Anthropic · Barclays scales Claude to upgrade operations and improve client experience — Anthropic

Claude Code opens itself to Mods

Version 2.1.287 of Claude Code adds Claude Mods, which means plugins may now modify deeper behavior. The release also includes a first built-in mod, “You should know”, in which a side agent watches over the session and flags things the user or Claude might miss. It is switched on with /plugin enable cc-plugin-you-should-know@builtin, and the release notes limit it to first-party sessions with telemetry on. Version 2.1.288 then gives mods a new interface element, $.ui.selection(). In fullscreen mode it returns the text the user last selected and, when the selection lies within a single transcript row, that row.

The three releases, 2.1.286 to 2.1.288, also carry smaller changes to everyday use. Permission prompts now show a count such as “2 of 5” when several requests stack up, and the “N more” rows of lists in fullscreen mode respond to the mouse. The agents view gains an n: filter that matches session names and tasks. Pressing Up on an empty prompt restores a draft cleared with Ctrl+C, including pasted text and images. Cloud sessions whose image lacks the GitHub CLI get a built-in gh api. That built-in no longer sends control characters from file names, jq filters or GitHub errors to the terminal. Another fix stops several Claude Code processes and IDE extensions from each opening a login browser when gcpAuthRefresh or awsAuthRefresh credentials expire.

Sources: Claude Code 2.1.287 — Added Claude Mods: plugins may now modify deeper behavior — Claude Code · Claude Code 2.1.288 — Added $.ui.selection() for mods — Claude Code · Claude Code 2.1.286 — Added a count such as "2 of 5" to the permission prompt — Claude Code

Sonnet 4.5 gets a retirement date

In its API release notes dated September 30, Anthropic announced the deprecation of Claude Sonnet 4.5 (claude-sonnet-4-5-20250929). Retirement on the Claude API is scheduled for November 30, 2026, less than two months after this edition. Anthropic recommends migrating to Claude Sonnet 5.5 and refers readers to its Model deprecations page for details. Applications pinned to the older model identifier will need to move before that date.

Sources: Claude API release notes — September 30, 2026: Claude Sonnet 4.5 deprecation — Anthropic (release notes)

Research items on robots and work

Anthropic's research page carries a piece titled “What work can robots do?”, and a separate Anthropic item asks “Can we predict the jobs robots will do?”. Both pose the question of which work robots can or will perform. A third item is titled “Claude-shaped science”. The material available for this edition gives only titles and no summaries or findings, so we do not characterize their conclusions here.

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