Dateline: June 5, 2026 | Next update: June 12, 2026
A consolidation and expansion week. No new model launched, but three significant platform stories dominated: Project Glasswing nearly quadrupled in size to approximately 200 partner organisations, with the EU joining and new critical-infrastructure sectors added; Claude Code shipped its largest single update of the post-Opus-4.8 era, with Opus 4.8 now the default model on Max, Team Premium, and the API, plus background jobs, auto-loading plugins, and enterprise version guardrails; and Code with Claude Tokyo ran June 5–6 with sessions livestreamed globally.
Claude / Anthropic
Project Glasswing — expands to ~200 organisations, EU joins
Project Glasswing — Anthropic's controlled-access programme giving vetted organisations access to Claude Mythos Preview for defensive cybersecurity work — has roughly quadrupled in two months. The original April cohort of ~50 partners grew by approximately 150 new organisations on June 2, following the EU's separate accession on June 1. The expansion deliberately widened the sector mix beyond the technology-heavy original cohort into power, water, healthcare, communications, and hardware, reflecting the industries most exposed to the vulnerabilities Mythos has been finding.
~150 new organisations added on June 2, bringing total Glasswing membership to approximately 200 across 15+ countries. EU joined as a bloc on June 1. New sectors added: power, water, healthcare, communications, and hardware — all critical infrastructure providers whose exposure extends beyond their home countries. Mythos Preview partners are now using the model beyond vulnerability discovery: writing patches, running pre-release security checks, penetration testing simulations, automating threat detection, and rebuilding legacy codebases in memory-safe languages. Anthropic is also releasing — on a request basis to trusted security teams outside Glasswing — the specialised vulnerability-discovery tools it built internally for the programme. Claude Security (built on Opus 4.8, available in Enterprise beta) remains the path for teams that do not qualify for Glasswing access.
Glasswing total: ~200 organisations | New cohort: ~150 organisations, 15+ countries | EU: joined June 1 | New sectors: power, water, healthcare, comms, hardware | Mythos Preview: not publicly available, Glasswing partners only | Internal tools: available on request to trusted security teams | Claude Security (Opus 4.8): Enterprise beta, separate from Glasswing | Anthropic committed $100M in model credits to the programme
Best for: Security teams, critical infrastructure operators, government bodies — apply via anthropic.com/glasswing
Claude Code — Opus 4.8 default, background jobs, auto-loading plugins
This is the largest Claude Code feature drop since Week 16. Opus 4.8 is now the default model on Max, Team Premium, Enterprise pay-as-you-go, and the direct Anthropic API. The update also introduces background job execution, auto-loading plugins from local skill directories, a lean system prompt as the new default for all modern models, and a range of developer workflow improvements.
Opus 4.8 is now the default model on Max, Team Premium, Enterprise pay-as-you-go, and the Anthropic API (requires v2.1.154+). Defaults to high effort; use /effort xhigh for the hardest tasks. Background jobs: in claude agents, prefix any shell command with ! to run it as a background process you can attach to and detach from — also available as claude --bg --exec 'pytest -x'. Auto-loading plugins: plugins placed in .claude/skills directories now load automatically with no marketplace install required; claude plugin init scaffolds a new plugin. New /reload-skills command re-scans skill directories without restarting the session. SessionStart hooks can return reloadSkills: true to make freshly installed skills available in the same session. Lean system prompt is now the default for all models except Haiku, Sonnet, and Opus 4.7 and earlier. Dynamic Workflows (Opus 4.8): manage background workflow runs with /workflows. Fast mode now defaults to Opus 4.8 at $10/$50 per MTok (2x standard rate, 2.5x speed); Opus 4.7 and 4.6 fast mode stays at $30/$150; Opus 4.6 fast mode is fully deprecated.
Opus 4.8 default: requires v2.1.154+ | Default effort: high; /effort xhigh for harder tasks | Background jobs: ! prefix in claude agents or --bg --exec flag | Auto-loading plugins: .claude/skills directory, no marketplace step | /reload-skills: re-scans without restart | SessionStart reloadSkills: true | Lean prompt: default for Opus 4.8, Sonnet 4.6, Haiku 4.5 | /workflows: manage Dynamic Workflow runs | Opus 4.6 fast mode: deprecated (CLAUDE_CODE_OPUS_4_6_FAST_MODE_OVERRIDE removed June 1)
Best for: All Claude Code users on Max, Team Premium, or API — update to v2.1.154+ to get Opus 4.8 as default
Claude Code — enterprise version guardrails and plugin management
Alongside the model and workflow updates, Claude Code shipped a focused enterprise management update: version guardrails allow IT and security teams to pin Claude Code to approved version ranges, and a new /plugin list command gives full visibility into installed plugins.
New managed settings: requiredMinimumVersion and requiredMaximumVersion — Claude Code refuses to start if its version falls outside the allowed range and directs the user to an approved version. Useful for enterprises that need to control which Claude Code version their fleet runs. New /plugin list command lists all installed plugins with --enabled and --disabled filters. New 'c to copy' shortcut in /btw copies the raw markdown answer to the clipboard with formatting preserved. Hooks improvement: Stop and SubagentStop hooks can now return hookSpecificOutput.additionalContext to give Claude feedback and continue the turn without being flagged as a hook error.
Managed settings: requiredMinimumVersion, requiredMaximumVersion (enterprise) | /plugin list: --enabled/--disabled filters | /btw copy shortcut: c | Hook return: hookSpecificOutput.additionalContext on Stop/SubagentStop | Fixed: EADDRINUSE errors when CLAUDE_CODE_TMPDIR is set to a deep path with tools binding Unix sockets | Fixed: terminal rendering performance (stabilised JIT compilation profile) | Fixed: large file write rendering performance | VSCode: tip added to disable terminal GPU acceleration to fix garbled glyphs (/terminal-setup)
Best for: Enterprise IT teams managing Claude Code fleet versions; developers using hooks and /btw
Code with Claude Tokyo — June 5–6
Code with Claude Tokyo runs today and tomorrow, completing the three-city 2026 developer event series (San Francisco May 6, London May 19–20, Tokyo June 5–6). Sessions cover building production-grade agents on the Claude Platform, Claude Code at scale (long-horizon tasks, multi-repo work, parallel agents, and surrounding infrastructure), and the roadmap from the Anthropic teams building these tools. Simultaneous interpretation is available in English and Japanese throughout.
Code with Claude Tokyo is live today (June 5). All Day 1 keynotes and breakout sessions are being livestreamed. Sessions include: what current models can do and where they're headed (from Anthropic researchers); production-grade agent design (with Asana, Cursor, GitHub, Replit, and Vercel); Claude Code at scale; and redesigning the development process for AI delegation. Office hours run all day alongside sessions. Register at claude.com/code-with-claude/tokyo for the livestream link.
Livestream: all Day 1 keynotes and breakout sessions | Language: primarily English, some Japanese sessions, live simultaneous interpretation both directions | Office hours: English only | Location: Tokyo (in-person applications closed; livestream open) | Series: SF (May 6), London (May 19–20), Tokyo (June 5–6)
Best for: Developers, engineering leads, and technical teams in APAC and globally — watch the livestream at claude.com/code-with-claude/tokyo
Claude Code security plugin — three-pass review model
The security-guidance plugin, which launched last week, received a detailed specification update clarifying exactly how its three-pass review model works.
Three-pass review architecture now documented: (1) fast pattern check runs on each code edit using regex across ~25 vulnerability classes; (2) model review runs at the end of each turn using Opus 4.8; (3) deeper agentic review triggers on commit or push. Project-level security rules can be added in .claude/claude-security-guidance.md to customise what the plugin flags. Internal data continues to show a 30–40% reduction in security-related PR comments since the plugin was introduced across teams using it.
Pass 1: regex pattern check per edit (~25 vulnerability classes) | Pass 2: Opus 4.8 model review per turn | Pass 3: agentic review on commit/push | Config: .claude/claude-security-guidance.md | Install: /plugin install security-guidance@claude-plugins-official | Complements: Claude Security (Enterprise, deeper codebase-wide scanning with data-flow tracing)
Best for: All Claude Code users wanting automated security feedback; enterprise teams wanting to customise rules per project
Model notes — Opus 4.7 remains available, no new model this week
No new model launched this week. Opus 4.8 (launched May 28) is the current flagship and is now rolling out as the default across paid tiers. Opus 4.7 remains available for users who want to stay on it explicitly — it is not deprecated. Sonnet 4.6 and Haiku 4.5 are unchanged. Claude Mythos Preview remains restricted to Project Glasswing partners only with no public release date.
Opus 4.8 is now the default on Max, Team Premium, Enterprise pay-as-you-go, and the API (v2.1.154+). Opus 4.7 remains available and is not deprecated. Sonnet 4 and original Opus 4 retirement is still on track for June 15 — migrate to Sonnet 4.6 and Opus 4.8 respectively.
Opus 4.8: default on Max/Team Premium/Enterprise PAYG/API from v2.1.154+ | Opus 4.7: available, not deprecated | Sonnet 4 + original Opus 4: retiring June 15, 2026 | Mythos Preview: Glasswing only, no public date | Model strings: claude-opus-4-8, claude-opus-4-7, claude-sonnet-4-6, claude-haiku-4-5-20251001
Best for: API developers — action required if still on Sonnet 4 or original Opus 4, retiring June 15
Plans and Pricing
No pricing changes this week. The fast mode restructuring from last week is now fully in effect: Opus 4.8 fast mode is $10/$50 per MTok (2x standard, 2.5x speed). Opus 4.7 and Opus 4.6 fast modes remain at $30/$150, with Opus 4.6 fast mode now deprecated via API.
Opus 4.8 standard: $5/$25 per MTok | Opus 4.8 fast mode: $10/$50 per MTok | Opus 4.7/4.6 fast mode: $30/$150 per MTok | Opus 4.6 fast mode: deprecated (use /model claude-opus-4-6[1m] then /fast if needed until June 15) | Sonnet 4.6: $3/$15 per MTok | Haiku 4.5: low-cost tier
ChatGPT / OpenAI
Dateline: June 5, 2026 | Next update: June 12, 2026
Over the past week, OpenAI has focused on strengthening ChatGPT's role as an AI operating system for work, improving collaboration features, and continuing to refine agentic capabilities across research, projects, and workflow execution.
GPT-5.3 Standard — default model
More reliable handling of complex workflows that involve multiple tools and sources of information.
Improved consistency when combining web searches, documents, memory, and user instructions within the same session.
Context ~128k | Output ~4k–8k | Improved cross-tool context retention
Best for: General use
GPT-5.3 Pro — high-reasoning model
Improved performance on long-form strategic and analytical tasks.
Better handling of large research projects requiring multiple stages of reasoning.
Context ~200k (est.) | Improved reasoning persistence across extended workflows
Best for: Deep analysis
GPT-5.3 Mini — fallback model
Improved efficiency in routing lightweight tasks away from larger models.
Context ~64k | Better task classification and model allocation
Best for: Quick tasks
Agent Mode
Improved coordination between planning, execution, and revision stages of complex tasks.
Better state management | Improved multi-stage workflow orchestration
Best for: Task delegation
Deep Research
Improved synthesis of large research corpora and conflicting information.
Enhanced evidence ranking | Better cross-source integration
Best for: Research
Memory & Projects
Improved project-level organisation and retrieval of relevant context.
Enhanced project segmentation | Reduced retrieval of irrelevant memories
Best for: Ongoing workflows
Advanced Voice Mode
More natural handling of follow-up questions and topic changes.
Improved conversational state tracking | Reduced context-loss during transitions
Best for: Voice interaction
Enterprise & Team Collaboration
Expanded support for collaborative workflows involving shared projects, documents, and team-based research.
Improved project architecture | Better collaboration layer integration
Best for: Organisational productivity
ChatGPT for Professional Workflows
Further improvements to structured outputs for research, reporting, planning, and documentation workflows.
Improved formatting consistency | Better workflow-specific optimisation
Best for: Knowledge work
Plans and Pricing
No significant pricing changes announced this week. Pricing stable | API structure unchanged.
Gemini (Google)
Date: June 5, 2026 | Next update: June 12, 2026
Following the massive wave of announcements at Google I/O, this week's focus shifted toward ecosystem migration schedules, foundational design restructuring in corporate tools, and external platform expansions.
Gemini 3.5 Flash — stable default
Adoption metrics indicate stable runtime across consumer endpoints. The model remains the standard default engine for the global rollout of the new AI Search Box interface.
Best for: High-volume production workloads
Gemini Omni Flash — creative scaling
Creative deployments continue scaling up across YouTube Shorts and Google Flow, following the post-I/O release of native cross-modal video editing capabilities.
Best for: Video and creative workflows
Gemini 3.5 Pro — restricted preview
Remains in restricted preview deployment phase ahead of the targeted public release scheduled for later in June.
Best for: Watch this space — no action needed yet
Google Chat interface redesign — breaking change
Google is restructuring how Gemini operates within enterprise workspaces. The current Gemini "side panel" will be completely abolished and replaced by a standalone conversational app called "Ask Gemini" starting the week of June 25, 2026.
The Gemini side panel in Google Chat will be replaced by a standalone "Ask Gemini" app interface from the week of June 25, 2026. Critical action required: past conversation history will not automatically migrate to the new "Ask Gemini" environment. Enterprise users must manually export necessary records before the June 25 deadline. System administrators are encouraged to issue workspace-wide alerts to prevent loss of local prompt configurations.
Best for: Enterprise Workspace admins — issue export alerts to all users before June 25
Apple Siri integration
Apple has finalised partnerships to leverage Google Gemini alongside NVIDIA infrastructure to revamp Siri's processing architecture for advanced context reasoning.
Best for: Apple device users — improved Siri reasoning capabilities coming
Android security patch — June 2026
The June 2026 Android security patch addresses 124 system vulnerabilities, including a critical privilege escalation exploit and a memory-corruption flaw affecting Qualcomm graphics drivers. Users utilising Gemini-integrated mobile devices are urged to apply immediately.
124 system vulnerabilities patched. Critical issues: privilege escalation exploit and memory-corruption flaw in Qualcomm graphics drivers. Apply immediately on all Gemini-integrated Android devices.
Best for: All Android users — apply the June patch immediately
Workspace — promotional access concluded
The promotional trial enabling standard Workspace accounts to use "Match Writing Style" and "Match Doc Format" officially expired on June 1, 2026. A paid plan is now required.
Best for: Workspace users still using these features — upgrade to a paid plan
Microsoft Copilot
Dateline: June 5, 2026 | Next update: June 12, 2026
The biggest story of the week is model choice and interface evolution: Microsoft introduced Anthropic Claude Opus 4.8 as a new model option inside Copilot, rolled out a major redesign of the Copilot app, and expanded capabilities across Word, PowerPoint, Outlook, Teams, SharePoint, and the mobile app.
Claude Opus 4.8 now available in Microsoft 365 Copilot
Microsoft added Anthropic Claude Opus 4.8 as a model option for advanced reasoning, long-running workflows, and complex multi-step tasks. Supports complex workflows, improved tool selection, stronger instruction following, and long-context document/presentation work. Integrated with Work IQ for organisation-grounded outputs.
Best for: Enterprise teams needing deeper reasoning and long-horizon task execution
Copilot app redesign — unified workspace and faster performance
Copilot now features a streamlined, chat-centred interface with improved navigation, project-organised Notebooks, and a unified workspace for context and actions.
Progressive disclosure UI | Left panel for agents, conversations, history | Task-aware prompt line | Work IQ–powered contextual actions | App loads 2x faster
Best for: Heavy Copilot users managing multi-document or multi-project workflows
Copilot Notebooks — major enhancements
Notebooks now sync across Copilot and OneNote, organise work by project, and unify chats, outputs, and references.
New UI for project-based organisation | Cross-app synchronisation | Improved research and drafting workflows
Best for: Researchers, analysts, students, multi-document creators
GPT-5.5 Instant — now powering Copilot Chat and Copilot Studio
GPT-5.5 Instant is now the default fast-response model for Copilot Chat and Studio.
Improved speed, summarisation, and reasoning | Optimised for interactive chat and agent workflows
Best for: Users needing fast, iterative drafting and Q&A
Outlook — voice, account selector, and smarter email workflows
Copilot in Outlook now supports voice input, an account selector, and improved grounding for email tasks.
Voice input | Multi-account Copilot context | Improved summarisation and drafting based on email threads
Best for: High-volume email users, executives, customer-facing roles
Word — document summaries, personalisation, writing suggestions
Copilot now generates document summaries, personalised suggestions, and improved writing guidance.
Enhanced grounding | Improved long-document comprehension | Better rewrite and tone controls
Best for: Writers, analysts, legal teams, proposal creators
PowerPoint — slide explanations and built-in skills
Copilot can now explain slides, interpret deck structure, and use built-in skills for layout and content refinement.
Slide-level reasoning | Improved narrative generation | Integration with Notebooks for deck creation
Best for: Presenters, sales teams, educators
Teams — call delegation, recap deletion, video recap
Copilot now supports call delegation, recap deletion, and video recap generation.
AI-assisted call handling | Privacy-aligned recap deletion | Video recap generation for meetings
Best for: Executives, assistants, customer support, meeting-heavy teams
SharePoint — custom Copilot skills
SharePoint now supports custom Copilot skills, enabling tailored content generation and site-specific automations.
Custom skill definitions | Integration with SharePoint content types | Enterprise governance support
Best for: Intranets, documentation teams, enterprise content managers
Mobile — refreshed design and Apple CarPlay
Copilot mobile app gains a refreshed design and Apple CarPlay support.
Streamlined chat UI | Improved navigation | CarPlay for hands-free interactions
Best for: Mobile-first professionals, field workers, commuters
Copilot controls — ISO certifications and watermarks
Expanded ISO certifications across the Copilot portfolio and optional watermarks for sensitive material. New Copilot Adoption Hub for admins.
Compliance expansion | Watermarking for sensitive outputs | New Copilot Adoption Hub for admins
Best for: Regulated industries, compliance teams, IT admins
Plans and Pricing
No pricing changes this week. Updates focus on model choice, UI redesign, and cross-app enhancements. Claude Opus 4.8 included for Copilot licensed users; GPT-5.5 Instant default in Chat/Studio; redesign rolling out globally.