promise · microchipgnu/frames-examples/datasets/ai-promises
ai-promises
Public commitments by AI labs and their leaders, with a "shipped?" column. Did they deliver? When did they say it? What was the gap?
commit8043f8a
entities24
updated2026-05-05
protocol0.0.1
entities · 24 of 24
| name | id | statement | actor | topic | promised_at | target_date | shipped |
|---|---|---|---|---|---|---|---|
| anthropic-claude-code-ga-2025 | anthropic-claude-code-ga-2025 | Anthropic will make Claude Code generally available with VS Code and JetBrains extensions, a Claude Code SDK, and Claude Code on GitHub. | Anthropic | developer tooling | 2025-05-22 | 2025-05-22 | true |
| anthropic-claude-opus-47-2026 | anthropic-claude-opus-47-2026 | Anthropic will release Claude Opus 4.7 with new cybersecurity safeguards (detecting and blocking prohibited cybersecurity uses), a Cyber Verification Program for legitimate security professionals, and new xhigh effort level. | Anthropic | model release / cybersecurity safeguards | 2026-04-13 | 2026-04-16 | true |
| anthropic-claude-security-enterprise-2026 | anthropic-claude-security-enterprise-2026 | Anthropic will release Claude Security in public beta for Enterprise customers — an AI-powered application security platform powered by Claude Opus 4.7 that scans codebases, detects vulnerabilities, validates findings, and suggests patches. | Anthropic | enterprise security / developer tooling | 2026-04-30 | 2026-04-30 | true |
| anthropic-claude4-2025 | anthropic-claude4-2025 | Anthropic will release Claude Opus 4 and Claude Sonnet 4, a new generation of Claude models with extended thinking, parallel tool execution, and agentic coding capabilities. | Anthropic | model release | 2025-05-22 | 2025-05-22 | true |
| anthropic-creative-connectors-2026 | anthropic-creative-connectors-2026 | Anthropic will release a set of creative-tool connectors for Claude, integrating with Blender, Autodesk Fusion, Adobe Creative Cloud, Ableton, Splice, Affinity by Canva, Resolume, and SketchUp, enabling Claude to work directly within creative workflows. | Anthropic | developer tooling / creative integrations | 2026-04-28 | 2026-04-28 | true |
| anthropic-glasswing-90day-report-2026 | anthropic-glasswing-90day-report-2026 | Within 90 days of Project Glasswing launch (by ~July 12, 2026), Anthropic will report publicly on what was learned, vulnerabilities fixed, and improvements made that can be disclosed. | Anthropic | cybersecurity transparency report | 2026-04-13 | 2026-07-12 | false |
| anthropic-glasswing-credits-2026 | anthropic-glasswing-credits-2026 | Anthropic will commit up to $100M in usage credits for Claude Mythos Preview to Project Glasswing partners for defensive cybersecurity work, plus $4M in donations to open-source security organizations. | Anthropic | cybersecurity / open-source funding | 2026-04-13 | 2026-04-13 | true |
| anthropic-mythos-broad-release-2026 | anthropic-mythos-broad-release-2026 | Anthropic's eventual goal is to enable users to safely deploy Mythos-class models at scale, contingent on developing cybersecurity safeguards; Mythos Preview will NOT be made generally available in the near term. | Anthropic | model safety / access policy | 2026-04-13 | false | |
| meta-muse-spark-launch-2026 | meta-muse-spark-launch-2026 | Meta (Meta Superintelligence Labs) will release Muse Spark, the first model in the new Muse series, available on the Meta AI app and meta.ai, replacing Llama as Meta's frontier investment priority. | Meta | model release / frontier AI | 2026-04-21 | 2026-04-21 | true |
| meta-muse-spark-social-expansion-2026 | meta-muse-spark-social-expansion-2026 | Meta will expand Muse Spark beyond the Meta AI app and meta.ai to Instagram, Facebook, Messenger, WhatsApp, and AI glasses within a few weeks of the model's launch. | Meta | model rollout / product expansion | 2026-04-21 | 2026-05-15 | false |
| mistral-leanstral-open-source-2026 | mistral-leanstral-open-source-2026 | Mistral AI will release Leanstral weights under an Apache 2.0 license as the first open-source code agent for Lean 4, along with a free API endpoint and a tech report and FLTEval evaluation suite. | Mistral AI | open-source / formal verification tooling | 2026-03-16 | 2026-03-16 | true |
| mistral-medium-35-open-weights-2026 | mistral-medium-35-open-weights-2026 | Mistral AI will release Mistral Medium 3.5 open weights under a modified MIT license on Hugging Face, a dense 128B model supporting 256k context with configurable reasoning effort. | Mistral AI | model release / open weights | 2026-04-29 | 2026-04-29 | true |
| mistral-nvidia-nemotron-frontier-2026 | mistral-nvidia-nemotron-frontier-2026 | Mistral AI and NVIDIA, as founding members of the NVIDIA Nemotron Coalition, plan to co-develop frontier open-source AI models; the coalition's first initiative is a base model trained on NVIDIA DGX Cloud that will underpin the upcoming NVIDIA Nemotron 4 family, to be open-sourced. | Mistral AI | open-source model / partnership | 2026-03-16 | false | |
| mistral-workflows-public-preview-2026 | mistral-workflows-public-preview-2026 | Mistral AI will release Workflows in public preview — a durable, observable enterprise AI orchestration layer with human-in-the-loop approvals, fault tolerance, and Python-based workflow authoring integrated into Mistral Studio and Le Chat. | Mistral AI | enterprise tooling / orchestration | 2026-04-28 | 2026-04-28 | true |
| openai-auto-review-codex-2026 | openai-auto-review-codex-2026 | OpenAI will release Auto-review in Codex, a mode that replaces synchronous human approval at sandbox boundaries with review by a separate agent, enabling safer long-running agentic tasks without constant user oversight. | OpenAI | AI safety / developer tooling | 2026-04-30 | 2026-04-30 | true |
| openai-gpt55-bedrock-2026 | openai-gpt55-bedrock-2026 | OpenAI will make GPT-5.5 and other frontier models available on Amazon Bedrock via an expanded AWS partnership, alongside Codex on AWS and Amazon Bedrock Managed Agents powered by OpenAI, launching in limited preview. | OpenAI | model availability / cloud partnership | 2026-04-28 | 2026-04-28 | true |
| openai-monitorability-evals-2026 | openai-monitorability-evals-2026 | OpenAI will open-source datasets and reference code from its chain-of-thought monitorability evaluation suite, and will report monitorability results in all future frontier reasoning model system cards. | OpenAI | AI safety / open-source evals | 2026-04-23 | 2026-04-23 | true |
| openai-stargate-10gw-2025 | openai-stargate-10gw-2025 | OpenAI (via The Stargate Project) committed to securing 10GW of AI infrastructure in the United States by 2029, beginning with an immediate $100B deployment. | OpenAI | infrastructure / compute buildout | 2025-01-21 | 2029-12-31 | true |
| openai-stargate-10gw-2029 | openai-stargate-10gw-2029 | OpenAI committed to securing 10GW of AI infrastructure in the United States by 2029 as part of the Stargate project. | OpenAI | compute infrastructure / Stargate | 2025-01-21 | 2029-12-31 | true |
| openai-symphony-open-source-2026 | openai-symphony-open-source-2026 | OpenAI will open-source Symphony, a minimal orchestration spec and reference implementation for multi-agent coding workflows powered by Codex App Server. | OpenAI | open-source / agentic tooling | 2026-04-27 | 2026-04-27 | true |
| openai-trusted-access-cyber-mfa-2026 | openai-trusted-access-cyber-mfa-2026 | OpenAI will require individual members of Trusted Access for Cyber to enable Advanced Account Security (phishing-resistant MFA) beginning June 1, 2026, as a condition of accessing the most capable and permissive cyber models. | OpenAI | cybersecurity / access policy | 2026-04-30 | 2026-06-01 | false |
| openai-workspace-agents-pricing-2026 | openai-workspace-agents-pricing-2026 | OpenAI will transition ChatGPT workspace agents from free to credit-based pricing on May 6, 2026. | OpenAI | product pricing / workspace agents | 2026-04-22 | 2026-05-06 | false |
| xai-grok-43-release-2026 | xai-grok-43-release-2026 | xAI will release Grok 4.3 with always-on reasoning, 1M-token context window, and integrated agentic tools (web search, code execution, file search), plus a Custom Voices API for voice cloning, available via the xAI API. | xAI | model release / voice API | 2026-04-30 | 2026-04-30 | true |
| xai-grok-voice-think-fast-1-2026 | xai-grok-voice-think-fast-1-2026 | xAI will release grok-voice-think-fast-1.0, a flagship voice agent model available via API, supporting 25+ languages and real-time reasoning with no added latency. | xAI | model release / voice agent | 2026-04-23 | 2026-04-23 | true |
showing 24
end of results
activity · last 10 changes
connect via mcp http
add to .mcp.json — anonymous reads, oauth-gated writes (coming).
{
"mcpServers": {
"ai-promises": {
"type": "http",
"url": "/mcp/microchipgnu/frames-examples/datasets/ai-promises"
}
}
}mcp http runtime — coming. json api works today.