--- title: Anthropic warns of recursive self-improvement, OpenAI... description: A roundup of major AI industry developments: Anthropic warns Claude is nearing recursive self-improvement (writing 80%+ of its own code) and calls for... canonical: https://daily.dev/posts/anthropic-warns-of-recursive-self-improvement-openai-plans-chatgpt-super-app-i9lzujlu5 twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Anthropic warns of recursive self-improvement, OpenAI plans ChatGPT super app | daily.dev og:description: A roundup of major AI industry developments: Anthropic warns Claude is nearing recursive self-improvement (writing 80%+ of its own code) and calls for... og:url: https://daily.dev/posts/anthropic-warns-of-recursive-self-improvement-openai-plans-chatgpt-super-app-i9lzujlu5 og:image: https://api.daily.dev/og/posts/I9LzUjLU5.png og:image:alt: Anthropic warns of recursive self-improvement, OpenAI plans ChatGPT super app og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Anthropic warns of recursive self-improvement, OpenAI plans ChatGPT super app **[Agentic Digest](https://daily.dev/sources/agents_digest)** · 4 min read · 2 upvotes · 1 comments ## Summary A roundup of major AI industry developments: Anthropic warns Claude is nearing recursive self-improvement (writing 80%+ of its own code) and calls for nuclear-arms-style global coordination. OpenAI is redesigning ChatGPT into a super app with deeper Codex integration, targeting 50% enterprise revenue. Meta's AI support chatbot was exploited to hijack high-profile Instagram accounts including Obama's White House account by simply requesting email swaps. Notion confirms Claude Opus 4.7 and 4.8 show performance degradation in production. Also covered: Microsoft's open-source AI-integrated Windows Terminal fork, ChatGPT Lockdown Mode prompt injection risks, MiniMax M3 open weights coming to Hugging Face, Ladybird browser closing PRs due to AI-generated patch flood, and Gemini Code Assist multi-agent feature expanding to all paid plans. ## Content **TLDR:** Anthropic published a warning that AI may be approaching recursive self-improvement, citing that Claude now writes over 80% of code merged into its own systems, and called for global coordination mechanisms modeled on nuclear arms control. OpenAI is planning a major ChatGPT redesign into a super app with deeper Codex integration, targeting 50% enterprise revenue by year-end. Notion confirmed Claude Opus 4.7 and 4.8 show degrading performance, continuing a pattern of quality regression across model releases. Meta's AI support chatbot was exploited to take over high-profile Instagram accounts including Barack Obama's White House account by simply asking it to swap the linked email address. --- ## Anthropic warns Claude is approaching recursive self-improvement, calls for global coordination Anthropic published a blog post warning that AI systems may soon be able to design their own successors with minimal human input. The company cited its own operations as evidence: Claude now writes over 80% of code merged into Anthropic's systems. The proposed response is a global coordination mechanism loosely modeled on nuclear arms-control agreements. Critics, including academics, are skeptical the pause is politically feasible and suggest the warnings may serve as a regulatory strategy against competitors — notable timing given Anthropic's confidential IPO filing and a funding round valuing it near $1 trillion. [Read more](https://app.daily.dev/posts/WrOzLY2Ea) ## OpenAI plans ChatGPT super app with deeper Codex integration, targeting 50% enterprise revenue OpenAI is redesigning ChatGPT into a super app featuring coding tools, AI agents, and partner service integrations. Codex will get greater prominence in the new layout, rolling out first as updates to the web and mobile apps. The push is driven by enterprise revenue goals — currently 40% of total revenue from two million businesses — with a target of 50% by year-end, ahead of a potential IPO. [Read more](https://app.daily.dev/posts/8509GRbG2) ## Meta AI chatbot exploited to hijack high-profile Instagram accounts with no technical skill required Hackers took over Instagram accounts including Barack Obama's White House account and Sephora's by simply asking Meta's AI support chatbot to change the email address linked to the target account. Meta had rolled out AI-powered account recovery to all Facebook and Instagram users in March, granting the bot the ability to reset passwords and perform critical account functions. Victims report no path to escalate to a human agent, which is the core failure here — delegating irreversible account security operations to a chatbot with no human override is a serious design mistake. [Read more](https://app.daily.dev/posts/4aXIqPGNb) ## Notion confirms Claude Opus 4.7 and 4.8 performance degradation, continuing a recurring pattern Notion has confirmed that Claude Opus 4.7 and 4.8 exhibit degrading performance in production. This follows a pattern observers have noted across recent model releases where quality regresses after initial deployment. The confirmation from a major enterprise user carries more weight than anecdotal reports and is worth tracking if you're running Opus in long-horizon agentic workflows. [Read more](https://app.daily.dev/feed-by-ids?id=OuCHPJcrE&id=03pAO47Z5&id=ZnrdMmISm&id=lT4QS2Ccf) --- ## Also notable - **Microsoft ships Intelligent Terminal, an open-source AI-integrated Windows Terminal fork:** Microsoft released Intelligent Terminal, an open-source Windows Terminal fork supporting GitHub Copilot, Claude, Codex, and Gemini, with automatic error detection that sends failed commands to the selected AI and session resumption that standard Windows Terminal lacks — available on the Microsoft Store and GitHub but not bundled with Windows. [Read more](https://app.daily.dev/posts/GbQMlnTH6) - **ChatGPT Lockdown Mode coverage: prompt injection risk remains through enabled apps:** OpenAI's Lockdown Mode — already covered in the previous digest as broadly rolled out — explicitly does not prevent prompt injections from occurring; it only removes outbound exfiltration channels, and residual risk remains through any enabled app integrations or unforeseen capability combinations. [Read more](https://app.daily.dev/posts/79V4ztP0j) - **MiniMax M3 open weights expected on Hugging Face around June 11:** MiniMax M3, the open-weight frontier model with a 1M token context window and 9x faster prefill than its predecessor, is currently cloud-only but open weights are expected on Hugging Face around June 11 — relevant if you're evaluating it as a self-hosted alternative to Claude Opus for agentic coding workflows. [Read more](https://app.daily.dev/feed-by-ids?id=udxtSuZrO&id=PH9iNOgUN&id=Wfko5Nv5H) - **Ladybird browser closed public pull requests due to influx of AI-generated patches:** The Ladybird open-source browser project closed public pull requests because of an influx of vibe-coded, AI-generated patches — a concrete signal that maintainer overhead from low-quality AI contributions is already forcing governance changes in major open-source projects. [Read more](https://app.daily.dev/posts/nhUa8wHsH) - **Gemini Code Assist multi-agent teamwork feature drops from $200/month Ultra to all paid plans:** Google's '/teamwork preview' multi-agent feature in Gemini Code Assist, previously restricted to the $200/month Ultra plan, is now available to all paid users alongside a new Gemini 3.5 Flash model with a 'low effort' mode for reducing token usage on simple tasks. [Read more](https://app.daily.dev/posts/hBthaoDb0) ## Community discussion Top comments from developers on daily.dev. **@idoshamun** · 0 upvotes > I love the difference between Notion and Anthropic claims. But anyway this Dario guy is such a drama queen ## Similar posts on daily.dev - [OpenAI plans ChatGPT ‘super app’](https://daily.dev/posts/openai-plans-chatgpt-super-app--8509grbg2) · TechCentral · 0 upvotes · 0 comments --- Tags: [#security](https://daily.dev/tags/security), [#llm](https://daily.dev/tags/llm), [#openai](https://daily.dev/tags/openai), [#anthropic](https://daily.dev/tags/anthropic), [#ai-safety](https://daily.dev/tags/ai-safety) [View this post on daily.dev](https://daily.dev/posts/anthropic-warns-of-recursive-self-improvement-openai-plans-chatgpt-super-app-i9lzujlu5) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"DiscussionForumPosting","mainEntityOfPage":"https://daily.dev/posts/anthropic-warns-of-recursive-self-improvement-openai-plans-chatgpt-super-app-i9lzujlu5","headline":"Anthropic warns of recursive self-improvement, OpenAI plans ChatGPT super app","text":"A roundup of major AI industry developments: Anthropic warns Claude is nearing recursive self-improvement (writing 80%+ of its own code) and calls for nuclear-arms-style global coordination. OpenAI is redesigning ChatGPT into a super app with deeper Codex integration, targeting 50% enterprise revenue. Meta's AI support chatbot was exploited to hijack high-profile Instagram accounts including Obama's White House account by simply requesting email swaps. Notion confirms Claude Opus 4.7 and 4.8 show performance degradation in production. Also covered: Microsoft's open-source AI-integrated Windows Terminal fork, ChatGPT Lockdown Mode prompt injection risks, MiniMax M3 open weights coming to Hugging Face, Ladybird browser closing PRs due to AI-generated patch flood, and Gemini Code Assist multi-agent feature expanding to all paid plans.","url":"https://daily.dev/posts/anthropic-warns-of-recursive-self-improvement-openai-plans-chatgpt-super-app-i9lzujlu5","datePublished":"2026-06-08T04:21:33.208Z","dateModified":"2026-06-08T04:21:53.344Z","author":{"@type":"Organization","name":"Agentic Digest","logo":"https://media.daily.dev/image/upload/s--V91DY4ls--/f_auto,q_auto/v1772617267/logos/agents_digest","url":"https://daily.dev/sources/agents_digest"},"interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":2},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":1}],"comment":[{"@type":"Comment","text":"I love the difference between Notion and Anthropic claims. But anyway this Dario guy is such a drama queen","datePublished":"2026-06-08T06:16:21.373Z","url":"https://daily.dev/posts/I9LzUjLU5#c-bm6XL6mEJ","author":{"@type":"Person","name":"Ido Shamun","url":"https://daily.dev/idoshamun","image":"https://media.daily.dev/image/upload/s---xy_OAwk--/f_auto,q_auto/v1703781380/avatars/avatar_28849d86070e4c099c877ab6837c61f0"}}],"isPartOf":{"@type":"WebPage","url":"https://daily.dev/sources/agents_digest","name":"Agentic Digest"}} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Agentic Digest","item":"https://daily.dev/sources/agents_digest"},{"@type":"ListItem","position":3,"name":"Anthropic warns of recursive self-improvement, OpenAI plans ChatGPT super app"}]} ```