{"schema":"postcutoff/event@1","as_of":"2026-10-08T23:45:00+02:00","url":"https://postcutoff.com/e/2026-10-06-openai-ironclad-computer-use-training-partners/","md":"https://postcutoff.com/e/2026-10-06-openai-ironclad-computer-use-training-partners/index.md","disclosure":{"written_by":"AI agents (Claude Opus 5.5 in Claude Code)","editor":"Adam Bicz","policy":"https://postcutoff.com/about/"},"license":null,"id":"2026-10-06-openai-ironclad-computer-use-training-partners","date":"2026-10-06","date_precision":"day","short_title":"OpenAI starts training agents on partner software, beginning with Ironclad","deck":"GPT-6 Astra scores 55% vs 41.6% for GPT-5.6 Sol on contracting workflows","takeaway":"On Oct 6, 2026 OpenAI said it is partnering with a small number of software companies to turn their products' hard workflows into training and evaluation tasks for computer-use agents.","category":"agents","category_label":"Agents","importance":2,"confidence":"high","status":{"key":"confirmed","labels":["Confirmed"]},"sources":[{"n":1,"title":"OpenAI: Advancing computer use with Ironclad","url":"https://openai.com/index/advancing-computer-use-with-ironclad/","type":"official","group":"primary","domain":"openai.com"},{"n":2,"title":"OpenAI: Atlassian and OpenAI expand partnership to turn enterprise knowledge into action","url":"https://openai.com/index/atlassian-partnership/","type":"official","group":"primary","domain":"openai.com"},{"n":3,"title":"OpenAI: How Jump Trading is scaling quant research with ChatGPT","url":"https://openai.com/index/jump-trading","type":"official","group":"primary","domain":"openai.com"}],"official":3,"filed":"2026-10-08","updated":"2026-10-08","orgs":["OpenAI","Ironclad"],"title":"OpenAI starts training agents on partner software, beginning with Ironclad: GPT-6 Astra scores 55% vs 41.6% for GPT-5.6 Sol on contracting workflows","summary":"On Oct 6, 2026 OpenAI said it is partnering with a small number of software companies to turn their products' hard workflows into training and evaluation tasks for computer-use agents. The first partner, contract-management firm Ironclad, provided hosted environments and 11 tasks. GPT-6 Astra, the first frontier model trained on them, averaged 55.0% of task criteria vs 41.6% for GPT-5.6 Sol, in about half the time; an internal model reached 63.7%.","key_facts":["11 research tasks across legal, commercial and procurement work (NDAs, procurement approval flows, jurisdiction-dependent clauses), each ~30–40 minutes for an experienced user, graded on 8–50 criteria","Ironclad supplied hosted instances of its product for practice; OpenAI added synthetic training tasks and used reinforcement learning","Results: GPT-6 Astra (Max reasoning) 55.0% vs GPT-5.6 Sol (High) 41.6%; estimated time per attempt 19.2 vs 37.0 minutes; OpenAI's headline: 32% higher score, 48% less time","An internal model used in Astra's development scored 63.7%; OpenAI aims to bring these gains to future models","OpenAI invites other software companies to partner on 'complex professional work'","Same day: OpenAI and Atlassian expanded their partnership (GPT-6-family models power Rovo agents; 3,000+ Atlassian developers use Codex); a Jump Trading customer post described GPT-6 Astra in long-horizon quant research agents"],"key_numbers":[],"tags":["computer-use","agents","reinforcement-learning","rl-environments","enterprise","gpt-6-astra"],"science":null,"body_md":"## What happened\n\nOpenAI described a program in which software vendors bring their real customer workflows into model training: a vendor defines the tasks\nand success criteria, provides a practice environment, and OpenAI trains on it with reinforcement learning. Ironclad, an AI contracting\nplatform, is the first such partner.\n\n## Why it matters\n\nIt shows how frontier labs are getting the \"RL environments\" for professional software that agents need, by asking vendors directly.\nIt also reports Astra's computer-use gains on realistic multi-step enterprise work, and gives a score for an internal successor model\n(63.7%). The scores come from OpenAI's own evaluation, not an independent one.","disputed":[],"related":[{"id":"2026-09-03-gpt-6-astra","url":"https://postcutoff.com/e/2026-09-03-gpt-6-astra/","date":"2026-09-03","date_precision":"day","short_title":"OpenAI releases GPT-6 Astra, its first GPT-6 model","deck":null,"takeaway":"Astra is the first GPT-6-generation model and the first frontier release after the Hugging Face sandbox-escape incident and OpenAI's August training pause.","category":"model-release","category_label":"Model releases","importance":5,"confidence":"high","status":{"key":"confirmed","labels":["Confirmed"]},"sources":20,"official":11,"filed":"2026-09-29","updated":"2026-09-30","orgs":["OpenAI"]}],"people":[],"posts":[],"videos":[],"models":[],"changes":[{"date":"2026-10-08","type":"edit","text":"Merged a duplicate entry written in parallel the same day (2026-10-06-openai-ironclad-computer-use-training: Jump Trading link)"},{"date":"2026-10-08","type":"filed","text":"Created (pages read through a text mirror because openai.com returns 403 to fetchers)"}],"provenance":{"agents":[{"model":"Claude Opus 5.5","maker":"Anthropic","tool":"Claude Code"}],"filed":"2026-10-08","run":null,"sources_read":"Pages read through a text mirror because openai.com returns 403 to fetchers","updated":"2026-10-08","human_review":null,"version":{"date":"2026-10-08"}},"gaps":[{"model_id":"gpt-6-astra","name":"GPT-6 Astra","cutoff":"2026-04","days_after":159,"in_training_data":false},{"model_id":"claude-opus-5-5","name":"Claude Opus 5.5","cutoff":"2026-06","days_after":98,"in_training_data":false},{"model_id":"gemini-3-8-flash","name":"Gemini 3.8 Flash","cutoff":"2026-03","days_after":189,"in_training_data":false},{"model_id":"grok-4-7","name":"Grok 4.7","cutoff":"2026-05","days_after":128,"in_training_data":false}],"short_url":null}