Here are today's news for 8th of October: Anthropic launched Claude Haiku 5.5 on October 7, 2026, positioning it as its fastest and cheapest small model for high, volume work such as classification, summaries, database queries and customer support; OpenAI announced GPT, 6 with “Intelligent UI” on October 7, 2026, beginning a global ChatGPT rollout for Plus, Pro, Business and Enterprise; Free and Go users were due to get access October 8; On October 6, Mistral AI opened a public, preview API for Mistral Large 4 (ML4), nicknamed “le Chonk,” and said its weights would follow by the end of October; On October 7, Meta brought its Muse AI assistant to iPad in a native app update, about a month after its September 8 launch on iPhone and Android; and On Sept. The first and most important one: Anthropic launched Claude Haiku 5.5 on October 7, 2026, positioning it as its fastest and cheapest small model for high, volume work such as classification, summaries, database queries and customer support. Let's dive into the details. From Anthropic: Anthropic launched Claude Haiku 5.5 on October 7, 2026, positioning it as its fastest and cheapest small model for high, volume work such as classification, summaries, database queries and customer support. It is available through Anthropic, Amazon Web Services, Google Cloud and Microsoft Azure under the API name `claude, haiku, 5, 5`. For prompts up to 100,000 tokens, input costs $0.10 and output $0.50 per million tokens; above that threshold, prices rise to $0.50 and $2.50. Anthropic says the model costs about 75% less to run than Haiku 4.5. Its 1, million, token context window, 128,000, token output limit and adjustable effort setting give developers room to trade reasoning depth against cost. On OSWorld 2.1 computer, use testing, Haiku 5.5 scored 72.4%, versus 15.7% for Haiku 4.5; on Terminal, Bench 4.0 it scored 39.2%. Those gains could make automated agents and high, volume enterprise services cheaper, while keeping harder work with Sonnet or Opus. The release completes Anthropic’s Claude 5.5 family, following Opus 5.5 and Sonnet 5.5. From OpenAI: OpenAI announced GPT, 6 with “Intelligent UI” on October 7, 2026, beginning a global ChatGPT rollout for Plus, Pro, Business and Enterprise; Free and Go users were due to get access October 8. Paid plans receive GPT, 6 Sol, while Free and Go receive GPT, 6 Luna. The feature lets ChatGPT combine text with charts, diagrams, buttons, forms and task, specific tools such as calculators, bill splitters and games, choosing a format to suit the request. OpenAI says a library of native, streamable components and a compiler render interfaces as the model generates them, so users can start reading before the full response is ready. This shifts competition beyond answer quality toward useful, interactive experiences inside the chat, and could affect users who now turn to separate apps for simple tools. Availability spans Instant through Extra High reasoning; GPT, 6 Astra’s Pro reasoning option does not support Intelligent UI. The change applies to Chat, not Work or Codex, and Enterprise access depends on administrator settings. OpenAI disclosed no pric From Mistral: On October 6, Mistral AI opened a public, preview API for Mistral Large 4 (ML4), nicknamed “le Chonk,” and said its weights would follow by the end of October. The company describes it as a natively multimodal, trillion, parameter model with 49 billion active parameters; developers can test it through Mistral Studio while the weights remain unreleased. Mistral says it will use the preview period to red, team the model with cybersecurity leaders, vetted partners and state authorities, then publish further architecture, benchmark and post, training details alongside the weights. The staged rollout makes the model available for evaluation before unrestricted deployment, while positioning Mistral to challenge US and Chinese open, weight rivals and appeal to organisations seeking alternatives to closed models. Mistral’s launch materials highlight coding, cybersecurity, finance, manufacturing and visual grounding, but provide no API price or full benchmark results yet. The release extends the French company’s push to build frontier models and Eu From Meta: On October 7, Meta brought its Muse AI assistant to iPad in a native app update, about a month after its September 8 launch on iPhone and Android. The free iPad app adapts Muse to the tablet’s larger screen and multitasking; the same update adds integrations with Canva, Dropbox, Figma, QuickBooks, GitHub, Zoom, Asana and Klaviyo, plus Granola and Notion. Muse helps with tasks such as email, meetings and reservations, and can connect to users’ accounts to carry out work. Meta says it seeks approval before taking action. The expansion follows an estimated 6.6 million installs across iOS and Android, according to Sensor Tower, and several weeks at No. 1 among free iPhone apps in the US. The rollout puts Meta’s agent in reach of tablet users and small businesses as tech firms compete to move beyond chatbots toward assistants that act across services. Meta had already added a Mac version shortly after the mobile debut; its rapid platform and connector expansion signals a push to make Muse a cross, device, connected assistant. From Google, Anthropic, Meta, OpenAI, xAI and NVIDIA: On Sept. 29, 2026, President Donald Trump and executives from Google, Anthropic, Meta, OpenAI, xAI and NVIDIA signed the one, page Joint Commitment on Frontier Responsibilities, a voluntary pledge for companies training or deploying frontier models. It calls for four layers of oversight: internal monitoring during training and deployment for capability and alignment risks, including cyber, biological and chemical threats, an internal team to check safeguards and fix problems, independent external audits, and an independent board committee to review reports and remediation. The companies also pledged to meet regularly on shared safety standards. The agreement matters because it puts the leading model developers and NVIDIA, a key supplier of AI computing, under a common framework, potentially shaping expectations across the industry. But it sets no audit method, implementation deadline, public, disclosure requirement or penalty; firms choose their evaluators, and results need not be published. Trump called the pact “morally binding,” not le That's all for today's AI news briefing. Have a great day!