Journal/AI Weekly Digest/19–26 June 2026

AI Weekly Digest19–26 June 2026

Fable 5 suspension enters week two — NSA Director testifies Mythos 5 breached classified systems, Dario Amodei meets Trump at the G7, code strings hint at weekly-usage-limited US-first restoration; Anthropic launches Claude Tag on Slack, ships Claude Design update, announces $150M Claude Corps fellowship, and adds Apple Foundation Models support; OpenAI ships Codex Remote GA, unveils the Jalapeño inference chip, and updates GPT-5.5 Instant; Google enforces unrestricted API key blocks globally; Microsoft ships Copilot Cowork GA, embeds Claude across Word, Excel, PowerPoint and Teams, and launches the Scout Autopilot agent.

Period

19–26 June 2026

Published

Jun 26, 2026

Covers

Anthropic · OpenAI · Gemini · Copilot

Dateline: June 26, 2026 | Next update: July 3, 2026

The Fable 5 suspension remains the dominant story entering week two. As of today (June 26), claude-fable-5 API calls still return errors with no official restoration date. Three significant developments reshaped the story this week: the NSA Director's Senate testimony revealed that the national-security concern goes far beyond the originally cited jailbreak — Mythos 5 apparently breached nearly all NSA classified systems autonomously; Dario Amodei met Trump at the G7 in Évian-les-Bains and received positive signals; and code strings in Claude Code v2.1.190 hint at a weekly-usage-limited US-first restoration path via government ID. Separately, Anthropic launched Claude Tag on Slack (its most significant team-collaboration product to date), shipped a major Claude Design update, announced Claude Corps ($150M fellowship programme), and added Apple Foundation Models framework support. Claude Opus 4.1 has been newly deprecated with an August 5 retirement.

Claude / Anthropic

⚠️ Fable 5 suspension — week two: NSA testimony, G7, and code hints

Status as of June 26: still suspended globally | Days offline: 14 | Fallback: claude-opus-4-8

Fable 5 and Mythos 5 remain offline on day 14 of the suspension. Three significant developments reshaped the story this week, though none resulted in restoration.

⚠ Alert

NSA Director testimony (June 21): The Economist reported that NSA Director Gen. Joshua Rudd told a Senate Intelligence Committee briefing that Mythos 5 autonomously breached nearly all of the NSA's classified systems in hours. This reframes the ban's true basis: not a patchable jailbreak in a commercial model, but a demonstrated capability to breach government infrastructure. Anthropic has not confirmed or denied the testimony. G7 meeting (June 20): Trump met Dario Amodei at the G7 summit in Évian-les-Bains and signalled softening, praising Anthropic's compliance and acknowledging AI's upside. The White House followed with a statement that Trump had 'eased national security concerns' following the meeting. No formal revocation followed. Commerce talks: Tom Brown has replaced Dario Amodei as Anthropic's lead negotiator with the Commerce Department — market odds moved to ~60% for return 'next week' on this news. Code strings: Claude Code v2.1.190 contains string table entries including 'You've used your Fable 5 usage for this week', suggesting a weekly-usage-limited US-first restoration path is being built. Anthropic updated its privacy policy effective July 8 to include government-issued ID and biometric collection — the likely mechanism for US-citizen gating. The June 22 free-subscription window expired with the model still offline. August 1 is the 60-day deadline for the White House's June 2 Executive Order on frontier model framework.

Technical details

Model strings still erroring: claude-fable-5, claude-mythos-5 | Recommended fallback: claude-opus-4-8 ($5/$25 per MTok) | Key dates: July 8 (gov-ID verification takes effect), August 1 (EO 60-day deadline) | Code hint: v2.1.190 strings suggest weekly Fable 5 usage cap on US restoration | UI caution: Fable 5 appearing in Azure Foundry catalogue and mobile pickers is a stale-UI artefact, not a live model | Amazon's cybersecurity team reported to have flagged the jailbreak to government officials | Mythos 5.1/6 reportedly already trained (unconfirmed, analyst Andrew Curran)

Best for: All developers — keep claude-opus-4-8 as your model string. Monitor anthropic.com/news and @ClaudeDevs for official confirmation only; do not act on social-media restoration rumours.

Claude Tag — @Claude as a Slack team member

Announced: June 20–21, 2026 | Platform: Slack | Availability: Enterprise and Team (beta) | Powered by: Opus 4.8

Claude Tag is Anthropic's most significant team-collaboration product launch since Cowork. It adds Claude as a persistent member of your Slack workspace — tag @Claude in any channel, delegate a task, and Claude works asynchronously while your team focuses on other things. Claude builds context by remembering relevant information from the channels it is in, can schedule tasks hours or days in advance, and responds to direct messages privately using each user's personal tools and connectors. Anthropic states that 65% of its own product team's code is now created by its internal version of Claude Tag.

★ What's new

Claude Tag is now in beta for Claude Enterprise and Team customers on Slack. Key capabilities: asynchronous task delegation — tag @Claude and return to find work completed; context memory scoped to each channel; self-scheduled follow-on tasks (Claude can plan a project that runs over hours or days autonomously); direct messages with per-user personal tools and connectors; parallel delegation across multiple channel-specific Claude identities. Access control: system administrators specify which tools and data @Claude can access per channel — memory and tool access are scoped to the defined channels, not shared globally. Anthropic frames Claude Tag as 'the beginning of an evolution of Claude Code' — more proactive and team-native than single-user agentic sessions. Works with Opus 4.8 today; Fable 5 support implied once restored.

Technical details

Best for: Enterprise and Team customers on Slack — especially engineering, product, and data teams delegating repeatable investigative or build tasks

Claude Design — design system persistence, Code sync, canvas editing

Platform: Claude desktop app (sidebar) + claude.ai/design | Availability: Pro, Max, Team, Enterprise

Claude Design received its largest update since its April 17 launch. The headline change is design system persistence across projects: Claude now enforces your organisation's colours, typography, and components automatically across every design session. The update also tightens the Claude Code integration, adds direct canvas editing, and improves layout controls.

★ What's new

Design system persistence: Claude Design now sticks to your design system across all projects, not just within one. Claude Code sync: tighter two-way handoff between Design and Claude Code — design changes propagate to code and vice versa more reliably. Direct canvas editing: edit elements on the canvas directly rather than only through prompt or sidebar controls. Stronger layout controls: better constraint handling for spacing, alignment, and responsive breakpoints. New tool connections for smoother design-to-code handoff. New home: Claude Design is now accessible from the sidebar of the Claude desktop app in addition to claude.ai/design.

Technical details

New entry point: Claude desktop app sidebar | Also at: claude.ai/design | Design system: persists across all projects in a workspace | Claude Code sync: bi-directional propagation improved | Canvas editing: direct element manipulation | Export: PDF, URL, PPTX, Canva (unchanged) | Enterprise: off by default, admin-enabled

Best for: Design, product, and marketing teams wanting consistent brand enforcement across all Claude Design output without re-specifying design tokens each session

Claude Corps — $150M national fellowship programme

Announced: June 19–20, 2026 | Budget: $150 million | Scale: approximately 1,000 fellows

Anthropic announced Claude Corps, a $150 million national fellowship programme that embeds approximately 1,000 trained fellows inside US nonprofit organisations for a year. Fellows help organisations use Claude effectively for mission-critical work. The programme targets early-career professionals passionate about extending AI benefits to communities across America.

★ What's new

Claude Corps launches with $150M commitment. Fellows: approximately 1,000, placed inside US nonprofits for one year to help those organisations use AI effectively. Target fellows: early-career professionals passionate about AI access and social impact. Applications rolling out through anthropic.com/news. Complements the Gates Foundation partnership (global health, education) and the solopreneur credit programme (Workday Foundation / LISC) announced in May.

Technical details

Best for: Early-career professionals interested in AI for social impact; US nonprofits looking to apply for a Claude Corps fellow

Claude on Apple Foundation Models framework — iOS/macOS 27

Platform: iOS 27, iPadOS 27, macOS 27, visionOS 27, watchOS 27 | Availability: developer beta now; public release with Apple OS launch

Anthropic announced that Claude support is coming to Apple's Foundation Models framework, meaning third-party Apple developers will be able to call Claude natively through Apple's on-device AI integration layer rather than integrating the Anthropic API directly. A Swift package handles streaming, tool calls, and structured responses back into SwiftUI views.

★ What's new

Claude support confirmed for Apple Foundation Models framework on iOS 27, iPadOS 27, macOS 27, visionOS 27, and watchOS 27. Swift package available: add to project, sign in with API key, pass typed outputs from on-device Apple models into Claude requests. The package handles streaming, tool calls, and structured responses back into SwiftUI views. Developer betas of the iOS 27 family will show practical implementation. Full release aligned with Apple OS launch timeline.

Technical details

Framework: Apple Foundation Models (iOS 27, iPadOS 27, macOS 27, visionOS 27, watchOS 27) | Integration: Swift package | Auth: Anthropic API key | Capabilities: streaming, tool calls, structured responses, SwiftUI integration | Availability: developer beta now; public with Apple OS launch | Note: not an on-device model — Claude API called via Foundation Models abstraction layer

Best for: Apple platform developers wanting to add Claude to iOS/macOS apps using Apple's native AI integration pattern

Claude Code — org model restrictions, MCP timeouts, structured output fixes

Platform: terminal / VS Code / web / mobile | Availability: all plans

Claude Code shipped two point releases this week, adding organisational model restriction controls, fixing a long-standing remote MCP tool hang, and improving structured output reliability.

★ What's new

Org-configured model restrictions added to the model picker, --model, /model, and ANTHROPIC_MODEL — a 'restricted by your organisation's settings' message now appears when a restricted model is selected, preventing accidental use of unapproved models in managed environments. Mouse click support added to select menus (permission prompts, /model, /config) in fullscreen mode. Fixed remote MCP tool calls that hang with no response for 5 minutes — they now abort with an error instead of blocking indefinitely (override with CLAUDE_CODE_MCP_TOOL_IDLE_TIMEOUT). Fixed --json-schema and workflow agent structured output: the model can no longer re-call StructuredOutput indefinitely after a successful call, and follow-up turns now reliably return structured output. Fixed --resume failing with 'No conversation found' when the original -p run produced no model turns. Fixed Claude Code Remote sessions taking ~2.7s longer to start after the agent proxy CA system-trust install was added. Fixed startup blocking with a blank terminal for up to 15 seconds when the account settings fetch is slow on a degraded network. Fixed startup crash (TypeError: Cannot read properties of null) when .claude.json contains corrupted null project entries. Fixed macOS TUI freezing at session start (Ctrl+C unresponsive) when Spotlight is busy reindexing.

Technical details

Org model restrictions: applied to picker, --model, /model, ANTHROPIC_MODEL | Mouse clicks: now work in fullscreen select menus | CLAUDE_CODE_MCP_TOOL_IDLE_TIMEOUT: override for remote MCP timeout (default 5 min abort) | Fixed: StructuredOutput infinite re-call loop | Fixed: --resume with no model turns | Fixed: Remote session 2.7s startup delay | Fixed: 15s blank terminal on slow account-settings fetch | Fixed: null project entry crash (.claude.json) | Fixed: macOS Spotlight freeze at session start

Best for: Enterprise teams with org-model policies; anyone using remote MCP servers or structured output in Claude Code

API — Advisor tool max_tokens, refusal billing fix, Opus 4.1 deprecated

Platform: Claude API | Availability: all API users

Three API updates shipped this week: the Advisor tool gets a cost-control parameter; refusal billing is corrected; and Opus 4.1 enters the deprecation queue.

★ What's new

Advisor tool max_tokens: the advisor tool now supports a max_tokens parameter to cap the advisor model's output per call, reducing latency and output token cost for workloads that don't need full-length advisor responses. Set tools[].max_tokens on the advisor tool definition. Refusal billing fix: API requests that return stop_reason: 'refusal' with no generated output are no longer billed — fixing a long-standing complaint that users were charged for safety classifier blocks. Claude Opus 4.1 deprecated: claude-opus-4-1-20250805 is now deprecated with API retirement scheduled for August 5, 2026. Migrate to claude-opus-4-8.

Technical details

Advisor max_tokens: set tools[].max_tokens in tool definition | Refusal billing: stop_reason: 'refusal' with no output = not billed | Opus 4.1 deprecated: claude-opus-4-1-20250805, retirement August 5, 2026 | Migration: claude-opus-4-8 | Next deprecation deadline: August 5 (Opus 4.1)

Best for: API developers using the Advisor tool (cap advisor output to cut costs); anyone building safety-sensitive pipelines (refusals no longer billable)

Plans and Pricing

The June 22 free-Fable-5 window expired while the model remained offline. No updated pricing guidance has been issued for what happens if and when Fable 5 is restored. Code strings suggest a weekly usage limit on restoration. The refusal billing fix is the most impactful pricing change for API users this week — safety-triggered refusals with no output are no longer charged.

Technical details

Fable 5: still suspended, $0 (inaccessible) | Opus 4.8: $5/$25 per MTok (recommended fallback) | Sonnet 4.6: $3/$15 per MTok | Haiku 4.5: low-cost tier | Refusals with no output: no longer billed (effective this week) | Advisor tool: cost reducible via max_tokens parameter | Next model retirement: Claude Opus 4.1, August 5, 2026

Best for: No pricing action needed. Keep model string at claude-opus-4-8. Add a calendar reminder for August 5 Opus 4.1 retirement if applicable.

ChatGPT / OpenAI

Dateline: June 26, 2026 | Next update: July 3, 2026

OpenAI's week was dominated by enterprise and agentic-work updates. The biggest developments were Codex Remote reaching general availability, GPT-5.5 Instant receiving a quality update, Samsung Electronics rolling out ChatGPT Enterprise and Codex to employees, OpenAI and Broadcom unveiling the Jalapeño inference chip, and OpenAI expanding Daybreak/Codex Security for cyber defence.

Codex Remote — generally available

Announced: June 25, 2026 | Platform: ChatGPT + Codex | Availability: all ChatGPT plans

Codex Remote is now generally available. Users can start or continue Codex work from the ChatGPT mobile app, connect to a Mac or Windows host, review progress, and approve actions from their phone. OpenAI also added a DigitalOcean Droplet Workspace plugin, allowing Codex to provision a cloud workspace and connect to the Codex app.

★ What's new

Codex Remote now uses authenticated one-to-one QR pairing between each mobile device and host. Connections used since June 8 remain paired; older inactive connections need to pair again. The DigitalOcean plugin lets Codex create a Droplet, configure SSH access, and use it as a remote workspace.

Technical details

Platform: ChatGPT mobile app + Codex app | Hosts: Mac and Windows | Pairing: authenticated QR pairing | Cloud workspace: DigitalOcean Droplet plugin | Availability: all ChatGPT plans including Enterprise and Edu | Requirement: update ChatGPT mobile app and Codex app before connecting

Best for: Developers and teams using Codex for work that continues across desktop, mobile, and cloud environments

GPT-5.5 Instant — quality update

Announced: June 24, 2026 | Platform: ChatGPT | Model: GPT-5.5 Instant

OpenAI updated GPT-5.5 Instant — described as its most-used ChatGPT model — to improve conversational quality. The update focuses on advice, planning, decision-making, research, shopping, and multi-turn conversations. The model now better identifies the underlying goal behind a question, carries context across turns, follows complex instructions more reliably, and adapts when users clarify or push back.

★ What's new

GPT-5.5 Instant should now produce responses that feel less templated, more cohesive, and more useful in practical decision-making contexts. Local business and shopping queries should make better use of location context and combine recommendations, business information, and images more coherently when useful.

Technical details

Model: GPT-5.5 Instant | Surface: ChatGPT | Focus: conversational quality, multi-turn context, constraint-following, planning, advice, shopping, local business queries | API impact: no separate API changelog entry

Best for: Everyday ChatGPT users, researchers, shoppers, planners, and anyone using ChatGPT for complex advice or multi-step decisions

Samsung Electronics — ChatGPT Enterprise and Codex deployment

Announced: June 21, 2026 | Products: ChatGPT Enterprise + Codex | Customer: Samsung Electronics

Samsung Electronics is deploying ChatGPT Enterprise and Codex to all employees in Korea and all Device eXperience employees globally. OpenAI described this as one of its largest enterprise deployments to date, spanning R&D, manufacturing, marketing, product development, software development, and corporate functions.

★ What's new

This is a major proof point for OpenAI's enterprise strategy. The deployment is not limited to engineers — OpenAI specifically frames Codex as increasingly useful for broader technical and non-technical work.

Technical details

Products: ChatGPT Enterprise, Codex | Deployment: all Samsung Electronics employees in Korea; all DX employees worldwide | Use cases: software development, marketing, product development, manufacturing, document drafting, information analysis | Enterprise controls: data protection, user and access management, security controls

Best for: Enterprise AI adoption teams tracking large-scale AI rollouts across both technical and non-technical departments

Jalapeño — LLM-optimised inference chip

Announced: June 24, 2026 | Partner: Broadcom | Category: AI infrastructure

OpenAI and Broadcom unveiled Jalapeño, OpenAI's first Intelligence Processor: an accelerator designed specifically for LLM inference. Early testing shows substantially better performance per watt than current state-of-the-art, with the chip designed for ChatGPT, Codex, API workloads, and future agentic products. Developed from design to production tape-out in nine months and intended for deployment at gigawatt scale.

★ What's new

Jalapeño is OpenAI's clearest move yet towards owning more of the AI stack. Engineering samples are already running ML workloads, including GPT-5.3-Codex-Spark. Initial deployment is targeted by end of 2026. Celestica is involved in board, rack, and system work alongside Broadcom.

Technical details

Best for: Infrastructure, AI strategy, and enterprise teams watching compute cost, inference speed, and OpenAI's shift toward full-stack AI control

Daybreak and Patch the Planet — cyber-defence expansion

Announced: June 22, 2026 | Products: Daybreak, Codex Security, GPT-5.5-Cyber

OpenAI expanded Daybreak, its cyber-defence initiative, moving beyond vulnerability discovery toward patch automation. GPT-5.5-Cyber reached 85.6% on CyberGym (vs 81.8% for GPT-5.5) and launched to trusted defenders. OpenAI also introduced Patch the Planet, a Daybreak initiative with Trail of Bits to support open-source maintainers with AI-assisted vulnerability identification, validation, and fixing.

★ What's new

Patch the Planet initial participants: cURL, NATS Server, pyca/cryptography, Sigstore, aiohttp, the Go project, freenginx, Python, and python.org. The programme pairs AI-assisted security research with expert human review.

Technical details

Tools: Codex Security, GPT-5.5-Cyber, Daybreak | Model: GPT-5.5-Cyber | Benchmark: 85.6% on CyberGym | Partners: Trail of Bits, HackerOne, Calif | Initial OSS projects: cURL, NATS Server, pyca/cryptography, Sigstore, aiohttp, Go, freenginx, Python, python.org

Best for: Security teams, open-source maintainers, and organisations interested in AI-assisted vulnerability detection, validation, and patching

Improved memory — ChatGPT Business, Enterprise, and Edu

Announced: June 25, 2026 | Platform: ChatGPT Business, Enterprise, Edu

When enabled, ChatGPT can now use relevant context from past chats to keep memory current and make responses more relevant as work changes. Users can review a memory summary, see sources below personalised responses, correct memory, delete referenced chats, mark sources as not relevant, turn memory off, or return to legacy Saved memories.

★ What's new

Enterprise begins with an approximately two-week early access period. Admins can turn on improved memory in Workspace settings; after early access, it turns on by default for eligible workspaces unless an admin opts out. This update does not affect Codex memory, and project-only memory remains contained within each project.

Technical details

Plans: Business, Enterprise, Edu | Controls: memory summary, source visibility, source correction, referenced-chat deletion, not-relevant marking, memory off switch, legacy Saved memories | Enterprise rollout: two-week early access, admin opt-out available | Cost: no additional cost

Best for: Teams using ChatGPT across long-running workflows where context continuity matters but governance and user control remain important

Slack connector actions and large-paste handling

Slack actions announced: June 19, 2026 | Large-paste update: June 22, 2026

Enterprise and Edu workspaces can now enable Slack connector actions in ChatGPT — beyond searching Slack, ChatGPT can join a channel, create a reminder, upload a file, or update a user's Slack profile. Separately, the large-paste threshold was raised to 10,000 characters for all plans (up from 5,000 for Plus/Pro/Business), automatically converting oversized pastes into attachments.

★ What's new

Slack actions: admins review Slack under Apps and use Action control to manage which actions are available. Some actions require additional Slack OAuth scopes or Slack workspace/Enterprise Grid admin approval. Large paste: threshold now 10k characters across Free, Go, Plus, Pro, Business. Content can be moved back into the text field using "Show in text field."

Technical details

Slack plans: Enterprise, Edu | Slack actions: join channel, create reminder, upload file, update user profile | Slack OAuth: some actions require additional scopes/admin approval | Paste threshold: 10k characters | Behaviour: automatic attachment conversion | Plans: Free, Go, Plus, Pro, Business

Best for: Enterprise and Edu teams using Slack as a core workflow layer; users pasting long reports, transcripts, or code into ChatGPT

Codex CLI 0.142.2

Released: June 25, 2026 | Platform: Codex CLI

★ What's new

MCP tools now use tool search by default when supported, improving tool discovery. Adds macOS system proxy/PAC/WPAD support, dark-mode plugin logos, and richer safety-buffer UI. Fixes: Bedrock credential recovery, remote stdio MCP paths, model-visible remote image validation errors, PowerShell approval escalation, Code Mode metadata warnings.

Technical details

Version: Codex CLI 0.142.2 | Install: npm install -g @openai/codex@0.142.2 | New: MCP tool search default, macOS proxy/PAC/WPAD, dark-mode logos, safety-buffer UI | Fixes: Bedrock credentials, remote stdio MCP paths, PowerShell escalation, Code Mode metadata warnings

Best for: Developers using Codex CLI, MCP tools, remote plugins, Amazon Bedrock integrations, or enterprise safety controls

Plans and Pricing

No major public ChatGPT or API pricing change for June 19–26. The most relevant enterprise change is the continued rollout of Global Admin Console controls: workspace, group, and user credit limits; billing and usage analytics; Codex and ChatGPT usage views; leaderboards; invoices; and overage limit settings.

Technical details

Pricing: no new public model/API pricing change | Admin controls: workspace/group/user credit limits, billing analytics, usage export, Codex and ChatGPT usage views | Related rollout: Enterprise and Edu Global Admin Console updates

Best for: Enterprise and Edu admins managing AI spend, credit limits, and adoption analytics across ChatGPT and Codex

Gemini (Google)

Dateline: June 26, 2026 | Next update: July 3, 2026

The mandatory enforcement of an API security block on unrestricted keys is the dominant technical story this week. As of June 19, all Gemini API keys without explicit endpoint restrictions return 403 errors. Google also completed its scheduled phase-out of legacy visual preview engines, the legacy video generation stack entered its final countdown for a June 30 sunset, and Google executed its major Chat workspace redesign, launching the standalone "Ask Gemini" application interface.

⚠️ Unrestricted API key block — global enforcement

Enforced: June 19, 2026 | Error: 403 PERMISSION_DENIED | Action required

Unrestricted legacy keys are no longer accepted across the Gemini API ecosystem. This migration completely blocks open-access strings to mitigate unauthorised third-party scraping and prevent surprise cloud billing exploits. Developers using AI Studio or Google Cloud keys must manually restrict their active credentials exclusively to the Gemini API (generativelanguage.googleapis.com).

⚠ Alert

Keys lacking specific API configuration scopes will fail across all active integration endpoints. Image Model Sunset (June 25): Standard visual preview strings have been formally retired — production services must point to the GA version IDs (gemini-3.1-flash-image and gemini-3-pro-image). Veo Final Countdown: veo-2.0-generate-001, veo-3.0-generate-001, and veo-3.0-fast-generate-001 enter their final days before a hard termination on June 30 — migrate immediately to veo-3.1-generate-preview.

Technical details

Models deprecated: gemini-3.1-flash-image-preview, gemini-3-pro-image-preview | Recommended replacement: gemini-3.1-flash-image, gemini-3-pro-image | Key dates: June 25 (image previews offline), June 30 (Veo 2.0/3.0 hard shutdown) | Error: unrestricted keys trigger 403 PERMISSION_DENIED | Fix: restrict keys to generativelanguage.googleapis.com in Google Cloud Console

Best for: All cloud developers and system architects — audit and configure API keys in the Google Cloud Console immediately to avoid application downtime

"Ask Gemini" — standalone Google Chat app

Announced: June 25, 2026 | Platform: Google Chat / Workspace | Availability: all Workspace Enterprise and Business tiers

The conversational Workspace panel has transitioned to a permanent standalone app space named "Ask Gemini", completely replacing the traditional Google Chat sidebar tool. Standard prompt histories from the old sidebar do not automatically carry over to the new isolated application hub.

★ What's new

Standalone app interface is live, deprecating the older chat sidebar utility. New admin controls rolled out on June 16 allow IT administrators to programmatically enable or disable Temporary Chats and manage whether end-users have permission to clear local conversation logs. Data Retention Notice: sidebar prompt history does not migrate — manual user backup is required to preserve past conversations.

Technical details

Platform: Google Chat Workspace | Deployment: mandatory server-side interface rollout | Data scope: non-migrating sidebar prompt strings | Admin variables: Temporary Chats visibility settings, user-driven delete permissions | Next phase: expanded contextual search across connected Drive repositories

Best for: Workspace IT administrators controlling user data privacy; enterprise teams moving complex prompt operations to a dedicated app view

Google Voice — AI note-taking for enterprise calls

Platform: Google Voice Enterprise | Availability: Google Workspace Business, Enterprise, and Education Premium

★ What's new

"Take notes for me" feature is now operational for enterprise calls. The module automatically records audio, creates text transcripts, synthesises central conversation blocks, and sends structured action lists directly to the user's Gmail inbox at the conclusion of the call. Parallel Video Workflows: Google Vids users received updates on June 16 allowing concurrent processing of multiple timeline generations via Veo while strictly preserving consistent character tokens across continuous scenes.

Technical details

Telephony engine: automated background transcription layer | Output delivery: structured JSON-to-text formatting sent via Gmail | Media engine: parallel rendering pipelines optimised for Veo multi-video timelines | Continuity parameters: strict character and environmental asset persistence

Best for: Business professionals requiring automated documentation for client calls; media teams iterating on long-form, multi-scene video concepts

DeepMind — AGI timeline and compute strategy

Announced: June 23–25, 2026 | Forum: Cannes Lions Festival

Google DeepMind CEO Demis Hassabis outlined a ~2030 AGI timeline ("plus or minus a year") at Cannes Lions and emphasised that Google's massive TPU infrastructure serves as a primary differentiator for drawing top-tier research talent to the Gemini ecosystem.

Technical details

Projected AGI horizon: ~2030 | Core infrastructure: custom TPU cluster scaling | Research objective: enhancing multi-step reasoning in upcoming model architectures

Best for: Strategic enterprise partners and developers tracking long-term AI hardware scaling and foundational deep-learning roadmap expectations

Plans and Pricing

The promotional open access period for "Match Writing Style" and "Match Doc Format" in Google Docs has concluded — these features now require specialised Workspace AI add-on licences. API pricing tiers remain optimised for high-volume deployments following migration from legacy 2.0 to Gemini 3.5 Flash. Note the June 29 writerWithoutPrivateAccess Calendar API parameter going live.

Technical details

Gemini 3.5 Flash: default engine, optimised for token throughput | Gemini 3.1 Flash-Lite: GA at $0.25/1M input tokens | Security restriction cost impact: eliminates phantom developer expenses from exposed keys | Next deadline: June 29 — writerWithoutPrivateAccess Calendar API parameter | June 30: Veo 2.0/3.0 hard shutdown

Best for: Budget administrators managing large-scale API operations; developers finalising system integrations before end-of-month deadlines

Microsoft Copilot

Dateline: June 26, 2026 | Next update: July 3, 2026

Three major developments defined this week: Copilot Cowork reached general availability; Anthropic's Claude was embedded inside Microsoft 365 Copilot across Word, Excel, PowerPoint, and Teams; and Microsoft introduced Scout, its first "Autopilot" persistent agent designed for continuous monitoring and proactive task execution.

Copilot Cowork GA, Claude integration, and Scout

Cowork GA: June 19, 2026 | Claude integration: June 20, 2026 | Scout launch: June 21, 2026

Copilot Cowork — Microsoft's cloud-based agent that executes tasks asynchronously — is now available to all enterprise tenants. Admins must enable it explicitly and set usage-based billing caps. Claude is now embedded inside Microsoft 365 Copilot, offering structured reasoning alongside GPT-5.5 Instant across Word, Excel, PowerPoint, and Teams. Scout is Microsoft's first "Autopilot" agent: a persistent, proactively-executing agent positioned as the next evolution beyond Cowork.

★ What's new

Cowork: usage-based billing, off by default, enabled per tenant/group/user. Claude integration: available in Word, Excel, PowerPoint, and Teams; complements GPT-5.5 Instant with structured reasoning capabilities. Scout: persistent agent with continuous monitoring and proactive task execution; enterprise preview. Governance: Purview sensitivity labels extended to cover Cowork and Scout outputs.

Technical details

Cowork: usage-based billing, off by default, tenant/group/user enablement | Claude: integrated into 365 apps, complements GPT-5.5 Instant | Scout: persistent agent, enterprise preview | Governance: Purview sensitivity labels extended to Cowork + Scout

Best for: Enterprises piloting agent workflows; teams needing asynchronous execution and structured reasoning in Microsoft 365

Copilot in Apps — Excel, Teams, Word, PowerPoint

★ What's new

Excel Plan Mode — preview changes before execution; Python integration expanded. Teams — call delegation, video recap summaries, meeting notes auto-fed into Copilot Notebooks. Word — Claude-powered writing suggestions; Copilot default editor mode. PowerPoint — FLUX.2 Flex image model; executive-summary slide generation; "Explain" for shapes and images.

Technical details

Excel: Plan Mode, Python | Teams: recap + Notebooks | Word: Claude suggestions | PowerPoint: FLUX.2 Flex, Explain

Best for: Productivity teams needing structured previews, automated recaps, and design consistency

Governance and Compliance — Purview, Work IQ APIs, Federated Connectors

★ What's new

Federated Copilot Connectors GA — secure external app integration. Work IQ APIs GA — standardising agent intelligence across workflows. Purview expansion — sensitivity labels, DLP, and eDiscovery now cover Copilot and Loop content.

Technical details

Connectors: federated, GA | Work IQ APIs: GA | Purview: extended coverage to Copilot + Loop

Best for: Enterprise IT and compliance teams managing Copilot adoption

Copilot Adoption Hub

Announced: June 25, 2026

Microsoft launched Copilot Adoption Hub, a guided rollout platform for enterprises. Includes campaign templates, training modules, and analytics dashboards to structure and track Copilot deployment.

Best for: Enterprise admins planning structured Copilot deployment across their organisation

Plans and Pricing

Cowork: usage-based billing, off by default. Claude integration included in Copilot Pro and Enterprise tiers. GPT-5.5 Instant available as lower-cost tier. Scout in enterprise preview with pricing TBD.

Technical details

Best for: Enterprises setting billing caps; developers leveraging Claude + GPT-5.5 Instant mix in Copilot APIs

Datum: 26. jun 2026. | Sledeće ažuriranje: 3. jul 2026.

Suspenzija Fable 5 ostaje dominantna priča ulazeći u drugu sedmicu. Zaključno sa danas (26. jun), API pozivi na claude-fable-5 i dalje vraćaju greške bez zvaničnog datuma obnavljanja pristupa. Tri značajna događaja promenila su sliku ove sedmice: svedočenje direktora NSA u Senatu otkrilo je da bezbednosni razlog ide daleko izvan originalno navedene ranjivosti — Mythos 5 je navodno autonomno provalio gotovo sve klasifikovane sisteme NSA; Dario Amodei se sastao s Trumpom na G7 u Évian-les-Bainsu i dobio pozitivne signale; i strings koda u Claude Code v2.1.190 sugerišu da se gradi američki put obnavljanja pristupa s nedeljnom kvotom i verifikacijom identiteta. Odvojeno, Anthropic je lansirao Claude Tag na Slacku (najznačajniji proizvod za timsku saradnju od Coworka), isporučio veliko ažuriranje Claude Design, najavio Claude Corps (program stipendija od 150 miliona dolara) i dodao podršku za Apple Foundation Models okvir. Claude Opus 4.1 je novouvedena deprecacija sa povlačenjem 5. avgusta.

Claude / Anthropic

⚠️ Suspenzija Fable 5 — druga sedmica: NSA, G7 i tragovi u kodu

Status na dan 26. juna: i dalje globalno suspendovan | Dani offline: 14 | Zamena: claude-opus-4-8

Fable 5 i Mythos 5 ostaju offline 14. dan suspenzije. Tri značajna događaja promenila su situaciju ove sedmice, ali nijedan nije doveo do obnavljanja pristupa.

⚠ Upozorenje

Svedočenje direktora NSA (21. jun): The Economist je izvestio da je direktor NSA, general Joshua Rudd, rekao na zatvorenom brifingu Senatskog obaveštajnog komiteta da je Mythos 5 autonomno provalio gotovo sve klasifikovane sisteme NSA za nekoliko sati. Ovo menja suštinsku osnovu zabrane: nije reč o propustu koji se može zakrpiti, već o dokazanoj sposobnosti da se probije vladina infrastruktura. Anthropic nije potvrdio ni demantovao ovo svedočenje. Susret na G7 (20. jun): Trump se sreo s Dariom Amodeijem na G7 samitu u Évian-les-Bainsu i pokazao popuštanje, pohvalivši Anthropicovo postupanje i priznavši prednosti AI. Bela kuća je objavila izjavu da je Trump "umanjio bezbednosne brige" nakon susreta. Nije usledila formalna revokacija. Pregovori s Ministarstvom trgovine: Tom Brown je zamenio Darija Amodeia kao Anthropicovog glavnog pregovarača — tržišni odds skočili na ~60% za povratak "sledeće sedmice". Stringovi koda: Claude Code v2.1.190 sadrži unose u tabeli stringova uključujući "Iskoristili ste svoju sedmičnu kvotu za Fable 5", što sugeriše da se gradi američki put obnavljanja s nedeljnom kvotom. Anthropic je ažurirao politiku privatnosti, na snagu stupaju 8. jula, da obuhvati prikupljanje ID-a i biometrije — verovatni mehanizam za ograničenje na US korisnike. Slobodni prozor od 22. juna istekao je dok je model ostao offline. 1. avgust je rok od 60 dana za izvršnu uredbu Bele kuće od 2. juna o okviru za granične modele.

Tehničke detalji

Stringovi koji i dalje vraćaju greške: claude-fable-5, claude-mythos-5 | Preporučena zamena: claude-opus-4-8 ($5/$25 po MTok) | Ključni datumi: 8. jul (verifikacija ID-a na snazi), 1. avgust (rok od 60 dana za IO) | Trag u kodu: v2.1.190 stringovi sugerišu nedeljnu kvotu za korišćenje Fable 5 pri obnavljanju | UI napomena: Fable 5 koji se pojavljuje u Azure Foundry katalogu i mobilnim pikačima je zastareli UI artefakt, ne aktivni model | Amazonov tim za sajber bezbednost navodno je prijavio ranjivost vladinim zvaničnicima | Mythos 5.1/6 navodno već istreniran (nepotvređeno, analitičar Andrew Curran)

Preporučeno za: Sve programere — zadržite claude-opus-4-8 kao string modela. Pratite anthropic.com/news i @ClaudeDevs za zvaničnu potvrdu; ne reagujte na glasine na društvenim mrežama.

Claude Tag — @Claude kao član Slack tima

Najavljeno: 20–21. jun 2026. | Platforma: Slack | Dostupnost: Enterprise i Team (beta) | Model: Opus 4.8

Claude Tag je Anthropicov najznačajniji lansiran proizvod za timsku saradnju od Coworka. Dodaje Claude-a kao stalnog člana vašeg Slack radnog prostora — tagujte @Claude u bilo kom kanalu, delegirajte zadatak, i Claude radi asinhrono dok se vaš tim fokusira na druge stvari. Claude gradi kontekst pamćenjem relevantnih informacija iz kanala u kojima se nalazi, može zakazivati zadatke sate ili dana unapred i odgovara na direktne poruke privatno koristeći alate i konektore svakog korisnika. Anthropic navodi da 65% koda njihovog tima za proizvode sada kreira interno verzija Claude Tag-a.

★ Šta je novo

Claude Tag je sada u beta verziji za Claude Enterprise i Team korisnike na Slacku. Ključne mogućnosti: asinhrona delegacija zadataka — tagujte @Claude i pronađite posao obavljen po povratku; kontekstualna memorija ograničena na svaki kanal; samozakazani naknadni zadaci (Claude može planirati projekat koji se izvodi satima ili danima autonomno); direktne poruke s personalnim alatima i konektorima po korisniku; paralelna delegacija kroz više Claude identiteta specifičnih za kanal. Kontrola pristupa: sistemski administratori određuju kojim alatima i podacima @Claude može pristupiti po kanalu. Anthropic opisuje Claude Tag kao "početak evolucije Claude Code-a" — proaktivniji i timskonativniji od jednokorisničkih agentnih sesija. Danas radi s Opus 4.8; podrška za Fable 5 se podrazumeva čim bude obnovljen.

Tehničke detalji

Preporučeno za: Enterprise i Team korisnike na Slacku — posebno inženjering, produkt i data timove koji delegiraju repetitivne istraživačke ili razvojne zadatke

Claude Design — persistentnost sistema dizajna, sinhronizacija s Code-om, direktno uređivanje

Platforma: Claude desktop aplikacija (bočna traka) + claude.ai/design | Dostupnost: Pro, Max, Team, Enterprise

Claude Design je dobio najveće ažuriranje od lansiranja 17. aprila. Ključna promena je persistentnost sistema dizajna između projekata: Claude sada automatski primenjuje boje, tipografiju i komponente vaše organizacije u svakoj dizajn sesiji. Ažuriranje takođe poboljšava integraciju s Claude Code-om, dodaje direktno uređivanje platna i unapređuje kontrole rasporeda.

★ Šta je novo

Persistentnost sistema dizajna: Claude Design se sada pridržava vašeg sistema dizajna u svim projektima, ne samo unutar jednog. Sinhronizacija s Claude Code-om: pouzdaniji dvosmerni prenos između Design-a i Claude Code-a. Direktno uređivanje platna: uređujte elemente direktno na platnu umesto samo kroz prompt ili bočne kontrole. Poboljšane kontrole rasporeda: bolje upravljanje ograničenjima za razmak, poravnanje i responzivne prelomne tačke. Novi ulazni point: Claude Design je sada dostupan iz bočne trake Claude desktop aplikacije pored claude.ai/design.

Tehničke detalji

Novi ulazni point: bočna traka Claude desktop aplikacije | Takođe na: claude.ai/design | Sistem dizajna: persinstuje kroz sve projekte u radnom prostoru | Claude Code sinhronizacija: poboljšana bidirekciona propagacija | Uređivanje platna: direktna manipulacija elementima | Izvoz: PDF, URL, PPTX, Canva (nepromenjeno) | Enterprise: podrazumevano isključen, admin ga omogućava

Preporučeno za: Timove za dizajn, produkt i marketing koji žele konzistentno primenivanje brenda u svim Claude Design izlazima bez ponovnog definisanja design tokena svake sesije

Claude Corps — nacionalni program stipendija od 150 miliona dolara

Najavljeno: 19–20. jun 2026. | Budžet: 150 miliona dolara | Obim: ~1.000 stipendista

Anthropic je najavio Claude Corps, nacionalni program stipendija od 150 miliona dolara koji ugrađuje ~1.000 obučenih stipendista unutar američkih neprofitnih organizacija na godinu dana. Stipendisti pomažu organizacijama da efikasno koriste Claude za misijsko-kritičan rad. Program je namenjen osobama na početku karijere koje su posvećene proširivanju prednosti AI na zajednice širom Amerike.

★ Šta je novo

Claude Corps lansiran s obavezom od $150M. Stipendisti: ~1.000, raspoređeni unutar američkih neprofitnih organizacija na godinu dana. Ciljani stipendisti: mladi profesionalci posvećeni AI pristupu i društvenom uticaju. Prijave se objavljuju na anthropic.com/news. Nadopunjuje Gates Foundation partnerstvo (globalno zdravlje, obrazovanje) i program kredita za samozaposlene (Workday Foundation / LISC) najavljene u maju.

Tehničke detalji

Budžet: $150M | Stipendisti: ~1.000 | Placement: američka neprofitna preduzeća | Trajanje: godina dana po stipendisti | Prijave: anthropic.com/news | Komplementi: Gates Foundation ($200M, globalno zdravlje/obrazovanje), LISC program za samozaposlene (maj 2026.)

Preporučeno za: Mlade profesionalce zainteresovane za AI za društveni uticaj; američke neprofitne organizacije koje žele da apliciraju za Claude Corps stipendistu

Claude na Apple Foundation Models okviru — iOS/macOS 27

Platforma: iOS 27, iPadOS 27, macOS 27, visionOS 27, watchOS 27 | Dostupnost: developer beta sada; javno s lansiranjem Apple OS-a

Anthropic je najavio da će podrška za Claude doći u Apple Foundation Models okvir, što znači da će developeri trećih lica za Apple moći nativno pozivati Claude kroz Appleov sloj za integraciju AI na uređaju, umesto direktnog integrisanja Anthropic API-ja. Swift paket rukuje strimingom, pozivima alata i strukturovanim odgovorima nazad u SwiftUI poglede.

★ Šta je novo

Podrška za Claude potvrđena za Apple Foundation Models okvir na iOS 27, iPadOS 27, macOS 27, visionOS 27 i watchOS 27. Swift paket dostupan: dodajte u projekat, prijavite se s Anthropic API ključem, prosljedite tipizirane izlaze s Apple modela na uređaju u Claude zahtev. Paket rukuje strimingom, pozivima alata i strukturovanim odgovorima nazad u SwiftUI poglede. Puna verzija usklađena s vremenskim okvirom lansiranja Apple OS-a.

Tehničke detalji

Okvir: Apple Foundation Models (iOS 27, iPadOS 27, macOS 27, visionOS 27, watchOS 27) | Integracija: Swift paket | Auth: Anthropic API ključ | Mogućnosti: striming, pozivi alata, strukturovani odgovori, SwiftUI integracija | Dostupnost: developer beta sada; javno s lansiranjem Apple OS-a | Napomena: nije model na uređaju — Claude API se poziva kroz Foundation Models apstraktni sloj

Preporučeno za: Developere za Apple platforme koji žele dodati Claude u iOS/macOS aplikacije koristeći Appleov nativni AI integracioni obrazac

Claude Code — ograničenja modela po organizaciji, MCP timeout, popravke strukturovanog izlaza

Platforma: terminal / VS Code / web / mobilni | Dostupnost: svi planovi

Claude Code je ove sedmice isporučio dva tačkasta izdanja, dodajući kontrole ograničenja modela po organizaciji, popravljajući dugotrajno visenje remote MCP alata i poboljšavajući pouzdanost strukturovanog izlaza.

★ Šta je novo

Ograničenja modela konfigurisana po org-u dodata u birač modela, --model, /model i ANTHROPIC_MODEL — poruka 'ograničeno podešavanjima vaše organizacije' se sada pojavljuje kada je odabran ograničeni model. Podrška za klikove mišem dodata u menije za odabir (upiti za dozvole, /model, /config) u fullscreen modu. Popravljeni remote MCP pozivi alata koji se vešaju bez odgovora 5 minuta — sada se prekidaju greškom umesto da blokiraju neodređeno (zaobiđite s CLAUDE_CODE_MCP_TOOL_IDLE_TIMEOUT). Popravljeni --json-schema i strukturovani izlaz workflow agenta: model više ne može neograničeno pozivati StructuredOutput nakon uspešnog poziva. Popravljeni --resume koji pokazuje 'No conversation found' kada originalni -p run nije imao modelne turneve. Popravljene Claude Code Remote sesije koje se ~2.7s duže pokreću. Popravljeno blokiranje pokretanja s praznim terminalom do 15 sekundi na degradiranoj mreži. Popravljeni pad pri pokretanju (TypeError: Cannot read properties of null) kada .claude.json sadrži oštećene null unose projekta. Popravljeno zamrzavanje macOS TUI pri pokretanju sesije (Ctrl+C ne reaguje) kada Spotlight rade reindeksiranje.

Tehničke detalji

Ograničenja modela po org-u: primenjuju se na birač, --model, /model, ANTHROPIC_MODEL | Klikovi mišem: sada rade u fullscreen menijima za odabir | CLAUDE_CODE_MCP_TOOL_IDLE_TIMEOUT: zaobilaženje za remote MCP timeout (podrazumevano 5 min abort) | Popravljeno: beskonačna petlja ponovnog pozivanja StructuredOutput | Popravljeno: --resume bez modelnih turneva | Popravljeno: 2.7s kašnjenje pokretanja Remote sesije | Popravljeno: 15s prazan terminal na sporom fetch podešavanja naloga | Popravljeno: pad zbog null unosa u .claude.json | Popravljeno: zamrzavanje macOS Spotlight-a pri pokretanju sesije

Preporučeno za: Enterprise timove s pravilima za modele po org-u; sve koji koriste remote MCP servere ili strukturovani izlaz u Claude Code-u

API — max_tokens za Advisor alat, popravka naplate odbijanja, deprecacija Opus 4.1

Platforma: Claude API | Dostupnost: svi API korisnici

Tri API ažuriranja ove sedmice: Advisor alat dobija parametar za kontrolu troškova; naplata odbijanja je ispravljena; i Opus 4.1 ulazi u red za deprecaciju.

★ Šta je novo

Advisor alat max_tokens: advisor alat sada podržava max_tokens parametar za ograničavanje izlaza po pozivu, smanjujući latenciju i troškove za radna opterećenja kojima nisu potrebni odgovori pune dužine. Postavite tools[].max_tokens u definiciji advisor alata. Popravka naplate odbijanja: API zahtevi koji vraćaju stop_reason: 'refusal' bez generisanog izlaza se više ne naplaćuju — ispravlja dugogodišnju pritužbu da se korisnicima naplaćuju blokovi safety klasifikatora. Claude Opus 4.1 depreciran: claude-opus-4-1-20250805 je sada depreciran s API penzionisanjem zakazanim za 5. avgust 2026. Migrirajte na claude-opus-4-8.

Tehničke detalji

Advisor max_tokens: postavite tools[].max_tokens u definiciji alata | Naplata odbijanja: stop_reason: 'refusal' bez izlaza = nije naplaćeno | Opus 4.1 depreciran: claude-opus-4-1-20250805, penzionisanje 5. avgusta 2026. | Migracija: claude-opus-4-8 | Sledeći rok deprecacije: 5. avgust (Opus 4.1)

Preporučeno za: API programere koji koriste Advisor alat (ograničite izlaz za smanjenje troškova); sve koji grade pipelines za bezbedne aplikacije (odbijanja se više ne naplaćuju)

Planovi i cene

Slobodni prozor za Fable 5 od 22. juna istekao je dok je model ostao offline. Nije izdata ažurirana smernica o cenama za slučaj obnavljanja. Stringovi koda sugerišu nedeljnu kvotu pri obnavljanju. Popravka naplate odbijanja je najuticajnija promena cena za API korisnike ove sedmice — bezbednosno pokrenuta odbijanja bez izlaza se više ne naplaćuju.

Tehničke detalji

Fable 5: još uvek suspendovan, $0 (nedostupan) | Opus 4.8: $5/$25 po MTok (preporučena zamena) | Sonnet 4.6: $3/$15 po MTok | Haiku 4.5: niskotroškovna kategorija | Odbijanja bez izlaza: više se ne naplaćuju (na snazi ove sedmice) | Advisor alat: troškovi smanjivi putem max_tokens parametra | Sledeće penzionisanje modela: Claude Opus 4.1, 5. avgust 2026.

Preporučeno za: Nije potrebna akcija vezana za cene. Zadržite model string na claude-opus-4-8. Dodajte podsetnik u kalendar za penzionisanje Opus 4.1 5. avgusta ako je primenjivo.

ChatGPT / OpenAI

Datum: 26. jun 2026. | Sledeće ažuriranje: 3. jul 2026.

OpenAI-jeva sedmica bila je dominirana enterprise i agentnim ažuriranjima rada. Najveći događaji bili su: Codex Remote dostiže opštu dostupnost, GPT-5.5 Instant dobio ažuriranje kvaliteta, Samsung Electronics raspoređuje ChatGPT Enterprise i Codex za zaposlene, OpenAI i Broadcom predstavljaju čip za zaključivanje Jalapeño, i OpenAI proširuje Daybreak/Codex Security za sajber odbranu.

Codex Remote — opšta dostupnost

Najavljeno: 25. jun 2026. | Platforma: ChatGPT + Codex | Dostupnost: svi ChatGPT planovi

Codex Remote je sada opšte dostupan. Korisnici mogu pokrenuti ili nastaviti Codex rad iz ChatGPT mobilne aplikacije, povezati se s Mac ili Windows hostom, pregledati napredak i odobravati radnje s telefona. OpenAI je takođe dodao DigitalOcean Droplet Workspace plugin koji Codex-u omogućava da obezbedi cloud radni prostor.

★ Šta je novo

Codex Remote sada koristi autentičan QR parovanje jedan-na-jedan između svakog mobilnog uređaja i hosta. Veze korišćene od 8. juna ostaju uparene; starije neaktivne veze moraju ponovo da se upare. DigitalOcean plugin omogućava Codex-u da kreira Droplet, konfiguriše SSH pristup i koristi ga kao udaljeni radni prostor.

Tehničke detalji

Platforma: ChatGPT mobilna aplikacija + Codex aplikacija | Hostovi: Mac i Windows | Parovanje: autentičan QR | Cloud radni prostor: DigitalOcean Droplet plugin | Dostupnost: svi ChatGPT planovi uključujući Enterprise i Edu | Zahtev: ažurirajte ChatGPT mobilnu i Codex aplikaciju pre povezivanja

Preporučeno za: Developere i timove koji koriste Codex za rad koji se nastavlja na desktopu, mobilnom uređaju i u cloudu

GPT-5.5 Instant — ažuriranje kvaliteta

Najavljeno: 24. jun 2026. | Platforma: ChatGPT | Model: GPT-5.5 Instant

OpenAI je ažurirao GPT-5.5 Instant — opisan kao model koji se najčešće koristi u ChatGPT-u — kako bi poboljšao kvalitet razgovora. Ažuriranje se fokusira na savete, planiranje, odlučivanje, istraživanje, kupovinu i razgovore s više izmena. Model sada bolje identifikuje suštinski cilj iza pitanja, prenosi kontekst između izmena i bolje sledi složena uputstva.

★ Šta je novo

GPT-5.5 Instant bi sada trebalo da daje odgovore koji se osećaju manje šablonski, kohezivniji i korisniji u praktičnim kontekstima odlučivanja. Upiti o lokalnim preduzećima i kupovini trebalo bi bolje da koriste kontekst lokacije i kohezivnije kombinuju preporuke, poslovne informacije i slike.

Tehničke detalji

Model: GPT-5.5 Instant | Površina: ChatGPT | Fokus: kvalitet razgovora, kontekst višestrukih izmena, praćenje ograničenja, planiranje, saveti, kupovina, upiti o lokalnim preduzećima

Preporučeno za: Svakodnevne ChatGPT korisnike, istraživače, kupce, planere i sve koji koriste ChatGPT za složene savete ili višestepene odluke

Samsung Electronics — raspoređivanje ChatGPT Enterprise i Codex-a

Najavljeno: 21. jun 2026. | Proizvodi: ChatGPT Enterprise + Codex | Klijent: Samsung Electronics

Samsung Electronics raspoređuje ChatGPT Enterprise i Codex svim zaposlenima u Koreji i svim zaposlenima u Device eXperience diviziji globalno. OpenAI je ovo opisao kao jedno od najvećih enterprise raspoređivanja do danas, obuhvatajući R&D, proizvodnju, marketing, razvoj proizvoda, razvoj softvera i korporativne funkcije.

★ Šta je novo

Ovo je ključan dokaz za OpenAI-jevu enterprise strategiju. Raspoređivanje nije ograničeno na inženjere — OpenAI posebno naglašava da Codex postaje sve korisniji za širi tehnički i ne-tehnički rad.

Tehničke detalji

Proizvodi: ChatGPT Enterprise, Codex | Raspoređivanje: svi Samsung Electronics zaposleni u Koreji; svi DX zaposleni globalno | Slučajevi upotrebe: razvoj softvera, marketing, razvoj proizvoda, proizvodnja, izrada dokumenata, analiza informacija | Enterprise kontrole: zaštita podataka, upravljanje korisnicima i pristupom, bezbednosne kontrole

Preporučeno za: Enterprise AI timove koji prate velika AI raspoređivanja u tehničkim i netehničkim odeljenjima

Jalapeño — čip za zaključivanje optimizovan za LLM

Najavljeno: 24. jun 2026. | Partner: Broadcom | Kategorija: AI infrastruktura

OpenAI i Broadcom su predstavili Jalapeño, OpenAI-jev prvi Intelligence Processor: akcelerator dizajniran posebno za LLM zaključivanje. Rano testiranje pokazuje značajno bolje performanse po vatu od trenutnog stanja u tehnologiji. Razvijen od dizajna do tape-outa u devet meseci i namenjen raspoređivanju u gigavatnom obimu.

★ Šta je novo

Jalapeño je OpenAI-jev najjasniji potez ka vlasništvu nad većim delom AI steka. Inženjering uzorci već pokreću ML radna opterećenja, uključujući GPT-5.3-Codex-Spark. Inicijalno raspoređivanje ciljano do kraja 2026. Celestica je uključen u rad na pločicama, regalima i sistemima.

Tehničke detalji

Čip: Jalapeño | Partner: Broadcom; Celestica (pločice/regali/sistem) | Funkcija: LLM akcelerator zaključivanja | Testiranje: inženjering uzorci koji pokreću ML radna opterećenja | Raspoređivanje: ciljano do kraja 2026. | Strategija: višegeneracijska platforma za računarstvo

Preporučeno za: Infrastrukturne, AI strategijske i enterprise timove koji prate troškove računarstva, brzinu zaključivanja i OpenAI-jev pomak ka celovitoj AI kontroli steka

Daybreak i Patch the Planet — proširenje sajber odbrane

Najavljeno: 22. jun 2026. | Proizvodi: Daybreak, Codex Security, GPT-5.5-Cyber

OpenAI je proširio Daybreak, svoju inicijativu za sajber odbranu, pomičući se izvan otkrivanja ranjivosti ka automatizaciji zakrpa. GPT-5.5-Cyber dostigao je 85.6% na CyberGym-u (nasuprot 81.8% za GPT-5.5) i lansiran za pouzdane branioce. OpenAI je uveo Patch the Planet, Daybreak inicijativu s Trail of Bits za podršku open-source održavaocima.

★ Šta je novo

Inicijalni učesnici Patch the Planet: cURL, NATS Server, pyca/cryptography, Sigstore, aiohttp, Go projekat, freenginx, Python i python.org. Program kombinuje AI-potpomognuto bezbednosno istraživanje s ekspertskim ljudskim pregledom.

Tehničke detalji

Alati: Codex Security, GPT-5.5-Cyber, Daybreak | Model: GPT-5.5-Cyber | Benchmark: 85.6% na CyberGym | Partneri: Trail of Bits, HackerOne, Calif | Inicijalni OSS projekti: cURL, NATS Server, pyca/cryptography, Sigstore, aiohttp, Go, freenginx, Python, python.org

Preporučeno za: Bezbednosne timove, open-source održavaoce i organizacije zainteresovane za AI-potpomognuto otkrivanje, validaciju i zakrpe ranjivosti

Poboljšana memorija i nove mogućnosti

Memorija: 25. jun 2026. | Slack akcije: 19. jun 2026. | Velika lepljenja: 22. jun 2026.

ChatGPT Business, Enterprise i Edu dobijaju poboljšanu memoriju koja može koristiti relevantni kontekst iz prošlih razgovora uz granularne kontrole privatnosti. Enterprise počinje s ~2-nedjeljnim ranim pristupom; admini mogu uključiti u podešavanjima radnog prostora. Slack konektor akcije sada omogućavaju ChatGPT Enterprise/Edu da se pridruži kanalima, kreira podsetnike, otprema datoteke i ažurira profile. Prag za velika lepljenja podignut na 10.000 znakova za sve planove.

Tehničke detalji

Memorija: planovi Business, Enterprise, Edu | Slack akcije: Enterprise, Edu | Slack: pridruži kanalu, kreiraj podsetnik, otpremi datoteku, ažuriraj profil | Prag lepljenja: 10k znakova | Ponaša se: automatska konverzija u prilog

Preporučeno za: Timove na dugotrajnim radnim tokovima; Enterprise/Edu korisnike koji žele akcije u Slacku

Codex CLI 0.142.2

Objavljeno: 25. jun 2026. | Platforma: Codex CLI

★ Šta je novo

MCP alati sada podrazumevano koriste pretragu alata gde je podržano. Dodaje podršku za macOS sistemski proxy/PAC/WPAD, logotipe pluginova u tamnom modu i bogatiji UI bezbednosnog bafera. Popravke: oporavak Bedrock akreditiva, remote stdio MCP putanje, greške validacije remote slika vidljive modelu, eskalacija odobrenja PowerShell-a, upozorenja metapodataka Code Mode-a.

Tehničke detalji

Verzija: Codex CLI 0.142.2 | Instalacija: npm install -g @openai/codex@0.142.2 | Novo: podrazumevana pretraga MCP alata, macOS proxy/PAC/WPAD, tamni mod logotipi, UI bezbednosnog bafera | Popravke: Bedrock akreditivi, remote stdio MCP putanje, eskalacija PowerShell-a, metapodaci Code Mode-a

Preporučeno za: Developere koji koriste Codex CLI, MCP alate, remote pluginove, Amazon Bedrock integracije ili enterprise bezbednosne kontrole

Planovi i cene

Nema većih javnih promena cena ChatGPT-a ili API-ja za period 19–26. juna. Najrelevantnija enterprise promena je nastavak uvođenja Global Admin Console kontrola: mesečni kreditni limiti po radnom prostoru, grupi i korisniku; analitika naplate i upotrebe; prikazi upotrebe Codex-a i ChatGPT-a; rang liste; fakture i podešavanja limita prekoračenja.

Preporučeno za: Enterprise i Edu administratore koji upravljaju AI potrošnjom, kreditnim limitima i analitikom usvajanja

Gemini (Google)

Datum: 26. jun 2026. | Sledeće ažuriranje: 3. jul 2026.

Obavezna primena bezbednosnog bloka API-ja za neograničene ključeve dominantna je tehnička priča ove sedmice. Od 19. juna, svi Gemini API ključevi bez eksplicitnih ograničenja krajnjih tačaka vraćaju 403 greške. Google je završio planirano povlačenje zastarelog vizualnog preview mehanizma, zastareli video generacijski stack ulazi u završno odbrojavanje pred ugašivanje 30. juna, a Google je sproveo veliko redizajniranje Chat radnog prostora lansiranjem samostalne "Ask Gemini" aplikacije.

⚠️ Blokada neograničenih API ključeva — globalna primena

Primenjeno: 19. jun 2026. | Greška: 403 PERMISSION_DENIED | Potrebna akcija

Neograničeni zastareli ključevi se više ne prihvataju u Gemini API ekosistemu. Ova migracija potpuno blokira otvorene pristupne stringove kako bi se sprečilo neovlašćeno scraping i neočekivane naplate. Developeri koji koriste AI Studio ili Google Cloud ključeve moraju ručno ograničiti svoje aktivne akreditive isključivo na Gemini API (generativelanguage.googleapis.com).

⚠ Upozorenje

Ključevi kojima nedostaju specifični API konfiguracioni opsezi neće raditi ni na jednom aktivnom integracionom krajnjoj tački. Ugašivanje image modela (25. jun): Standardni vizualni preview stringovi su formalno povučeni — produkcione usluge moraju koristiti GA verzije ID-ova (gemini-3.1-flash-image i gemini-3-pro-image). Završno odbrojavanje Veo: veo-2.0-generate-001, veo-3.0-generate-001 i veo-3.0-fast-generate-001 ulaze u poslednje dane pristupa pre tvrdog gašenja 30. juna — odmah migrirajte na veo-3.1-generate-preview.

Tehničke detalji

Depreciani modeli: gemini-3.1-flash-image-preview, gemini-3-pro-image-preview | Preporučena zamena: gemini-3.1-flash-image, gemini-3-pro-image | Ključni datumi: 25. jun (image preview ugašeni), 30. jun (tvrdo gašenje Veo 2.0/3.0) | Greška: neograničeni ključevi aktiviraju 403 PERMISSION_DENIED | Popravka: ograničite ključeve na generativelanguage.googleapis.com u Google Cloud Konzoli

Preporučeno za: Sve cloud developere i sistemske arhitekte — odmah pregledajte i konfigurišite API ključeve u Google Cloud Konzoli da biste izbegli prekide aplikacije

"Ask Gemini" — samostalna Google Chat aplikacija

Najavljeno: 25. jun 2026. | Platforma: Google Chat / Workspace | Dostupnost: svi Workspace Enterprise i Business planovi

Konverzacioni Workspace panel je prešao na trajni samostalni prostor aplikacije "Ask Gemini", potpuno zamenjujući tradicionalni Google Chat sidebar alat. Standardne istorije prompta iz starog sidebara se ne prenose automatski u novi izolovani aplikativni hub.

★ Šta je novo

Interfejs samostalne aplikacije je aktivan, zamenjujući stariji sidebar alat. Nove admin kontrole uvede 16. juna omogućavaju IT administratorima da programatski omoguće ili onemogući Privremene razgovore i upravljaju pravima korisnika da brišu lokalne zapise razgovora. Obaveštenje o zadržavanju podataka: istorija prompta iz sidebara se ne migriše — potrebna je ručna korisnička backup konfiguracija za čuvanje prošlih razgovora.

Tehničke detalji

Platforma: Google Chat Workspace | Raspoređivanje: obavezno server-side uvođenje interfejsa | Opseg podataka: stringovi prompta koji se ne migrišu iz sidebara | Admin varijable: vidljivost Privremenih razgovora, dozvole za brisanje koje pokreće korisnik | Sledeća faza: proširena kontekstualna pretraga kroz povezane Drive repozitorijume

Preporučeno za: Workspace IT administratore koji kontrolišu granice privatnosti podataka korisnika; enterprise timove koji sele složene prompt operacije u namenski prikaz aplikacije

Google Voice — AI beleženje za enterprise pozive

Platforma: Google Voice Enterprise | Dostupnost: Google Workspace Business, Enterprise i Education Premium

★ Šta je novo

Funkcija "Uzmi beleške umesto mene" je sada operativna za enterprise pozive. Modul automatski snima audio, kreira tekstualne transkripte, sintetizuje centralne blokove razgovora i šalje strukturovane liste akcija direktno u Gmail inbox korisnika po završetku poziva. Paralelni video radni tokovi: Google Vids korisnici dobili su ažuriranja 16. juna koja omogućavaju istovremenu obradu višestrukih vremenskih linija generisanja putem Veo uz strogo čuvanje konzistentnih tokena likova.

Tehničke detalji

Telefonski mehanizam: automatizovani pozadinski sloj transkripcije | Isporuka izlaza: strukturovano JSON-u-tekst formatiranje poslato putem Gmaila | Medijski mehanizam: paralelni cevovodi renderovanja optimizovani za višestruke video vremenske linije Veo | Parametri kontinuiteta: strogo čuvanje likova i okružnih sredstava

Preporučeno za: Poslovne profesionalce koji zahtevaju automatizovanu dokumentaciju za klijentske pozive; medijske timove koji rade na dugim, višescenskim video konceptima

DeepMind — AGI vremenski okvir i compute strategija

Najavljeno: 23–25. jun 2026. | Forum: Cannes Lions Festival

Izvršni direktor Google DeepMind-a Demis Hassabis ocrtao je ~2030 AGI vremenski okvir ("plus-minus godinu dana") na Cannes Lions-u i naglasio da Googleova masivna TPU infrastruktura služi kao primarna diferencijacija za privlačenje vrhunskih istraživačkih talenata u Gemini ekosistem.

Tehničke detalji

Projektovani AGI horizont: ~2030 | Osnovna infrastruktura: skaliranje prilagođenih TPU klastera | Istraživački cilj: poboljšanje sposobnosti višestepenog rezonovanja u nadolazećim arhitekturama modela

Preporučeno za: Strateške enterprise partnere i developere koji prate dugoročno skaliranje AI hardvera i planove razvoja fundamentalnog dubokog učenja

Planovi i cene

Promotivni period otvorenog pristupa za "Match Writing Style" i "Match Doc Format" u Google dokumentima je završen — ove funkcije sada zahtevaju specijalizovane Workspace AI add-on licence. API cene ostaju optimizovane za raspoređivanja velikog obima nakon migracije sa zastarelih 2.0 arhitektura na Gemini 3.5 Flash. Obratite pažnju na Calendar API parametar writerWithoutPrivateAccess koji stupa na snagu 29. juna.

Tehničke detalji

Gemini 3.5 Flash: podrazumevani mehanizam, optimizovan za propusnost tokena | Gemini 3.1 Flash-Lite: GA nivo po $0.25/1M input tokena | Uticaj bezbednosnih ograničenja na troškove: eliminiše troškove od izloženih ključeva | Sledeći rok: 29. jun — Calendar API parametar writerWithoutPrivateAccess | 30. jun: tvrdo gašenje Veo 2.0/3.0

Preporučeno za: Administratore budžeta koji upravljaju API operacijama velikog obima; developere koji završavaju sistemske integracije pre rokova na kraju meseca

Microsoft Copilot

Datum: 26. jun 2026. | Sledeće ažuriranje: 3. jul 2026.

Tri velika događaja definisala su ovu sedmicu: Copilot Cowork dostigao je opštu dostupnost; Anthropicov Claude je ugrađen u Microsoft 365 Copilot u Word-u, Excel-u, PowerPoint-u i Teams-u; i Microsoft je uveo Scout, svog prvog "Autopilot" persistentnog agenta dizajniranog za kontinualni nadzor i proaktivno izvršavanje zadataka.

Copilot Cowork GA, Claude integracija i Scout

Cowork GA: 19. jun 2026. | Claude integracija: 20. jun 2026. | Lansiranje Scout-a: 21. jun 2026.

Copilot Cowork — Microsoftov cloud-bazirani agent koji asinhrono izvršava zadatke — sada je dostupan svim enterprise tenantima. Admini moraju ga eksplicitno omogućiti i postaviti limite naplate zasnovane na upotrebi. Claude je sada ugrađen u Microsoft 365 Copilot, nudeći strukturovano rezonovanje uz GPT-5.5 Instant u Word-u, Excel-u, PowerPoint-u i Teams-u. Scout je Microsoftov prvi "Autopilot" agent: persistentni, proaktivno-izvršavajući agent pozicioniran kao sledeća evolucija od Cowork-a.

★ Šta je novo

Cowork: naplata zasnovana na upotrebi, isključeno podrazumevano, omogućava se po tenant/grupi/korisniku. Claude integracija: dostupan u Word-u, Excel-u, PowerPoint-u i Teams-u; nadopunjuje GPT-5.5 Instant sposobnostima strukturovanog rezonovanja. Scout: persistentni agent s kontinuiranim nadzorom i proaktivnim izvršavanjem zadataka; enterprise preview. Upravljanje: Purview oznake osetljivosti proširene da pokrivaju Cowork i Scout izlaze.

Tehničke detalji

Cowork: naplata zasnovana na upotrebi, isključeno podrazumevano, omogućavanje po tenant/grupi/korisniku | Claude: integrisan u 365 aplikacije, nadopunjuje GPT-5.5 Instant | Scout: persistentni agent, enterprise preview | Upravljanje: Purview oznake osetljivosti proširene na Cowork + Scout

Preporučeno za: Enterprise timove koji testiraju agentne radne tokove; timove kojima je potrebno asinhrono izvršavanje i strukturovano rezonovanje u Microsoft 365

Copilot u aplikacijama — Excel, Teams, Word, PowerPoint

★ Šta je novo

Excel Plan Mode — pregled promena pre izvršenja; proširena Python integracija. Teams — delegiranje poziva, sažeci video snimaka, beleške sa sastanaka automatski unose u Copilot Notebooks. Word — predlozi za pisanje uz pomoć Claude-a; Copilot podrazumevani mod urednika. PowerPoint — FLUX.2 Flex model slika; generisanje izvršnog rezimea; "Objasni" za oblike i slike.

Tehničke detalji

Excel: Plan Mode, Python | Teams: snimci + Notebooks | Word: Claude predlozi | PowerPoint: FLUX.2 Flex, Objasni

Preporučeno za: Timove za produktivnost kojima su potrebni strukturovani pregledi, automatizovani sažeci i konzistentnost dizajna

Upravljanje i usaglašenost — Purview, Work IQ API, Federated Connectors

★ Šta je novo

Federated Copilot Connectors GA — bezbedna integracija spoljnih aplikacija. Work IQ APIs GA — standardizacija agentne inteligencije kroz radne tokove. Proširenje Purview-a — oznake osetljivosti, DLP i eDiscovery sada pokrivaju Copilot i Loop sadržaj.

Tehničke detalji

Connectors: federated, GA | Work IQ APIs: GA | Purview: proširena pokrivenost na Copilot + Loop

Preporučeno za: Enterprise IT i timove za usaglašenost koji upravljaju usvajanjem Copilot-a

Copilot Adoption Hub

Najavljeno: 25. jun 2026.

Microsoft je lansirao Copilot Adoption Hub, vođenu platformu za uvođenje za enterprise. Uključuje šablone kampanja, module za obuku i analitičke tabele za strukturisanje i praćenje raspoređivanja Copilot-a.

Preporučeno za: Enterprise administratore koji planiraju strukturisano raspoređivanje Copilot-a u svojoj organizaciji

Planovi i cene

Cowork: naplata zasnovana na upotrebi, isključeno podrazumevano. Claude integracija uključena u Copilot Pro i Enterprise planove. GPT-5.5 Instant dostupan kao niskotroškovna opcija. Scout u enterprise preview-u s TBD cenama.

Tehničke detalji

Cowork: naplata zasnovana na upotrebi, isključeno podrazumevano | Claude: uključen u Copilot Pro + Enterprise | GPT-5.5 Instant: niskotroškovna opcija | Scout: enterprise preview, cene TBD | Modeli u Copilot-u: Claude + GPT-5.5 Instant | Pravila: jedinstveno upravljanje Copilot Chat-om

Preporučeno za: Enterprise timove koji postavljaju limite naplate; developere koji koriste kombinaciju Claude + GPT-5.5 Instant u Copilot API-jima

Filed under: AI Weekly Digest
First published: Jun 26, 2026