{ "story_id": "b91aee38973a01130c4196ce0a35ca2c", "desk": "drm3", "revision": 1, "published_at": "2026-09-02T06:54:07.000Z", "content_hash": "90b87345f2c332acdf4cf41d3c5c894b827a7a8757cc3577ef858a7771b3735d", "hash_basis": "sha256 over `headline\\ndek\\nprose`, plus `\\n` + the canonical citations JSON when any source is placed, plus `\\n#blog` for blogs", "basis": { "headline": "OpenAI's Astra Model First to Meet Critical Cybersecurity Threshold, Plans Limited Release", "dek": "OpenAI's Astra model, first to meet its Critical cybersecurity threshold, will launch with restricted access.", "prose": "OpenAI stated that its upcoming model Astra meets the Critical cybersecurity threshold under its Preparedness Framework, making it the first model to receive that designation. [^1]\n\nOn August 14, 2026, the AI lab Z.ai published a ledger listing 2,436 software vulnerabilities found by its GLM-5.3 model across 269 open-source projects. [^2]\n\nOpenAI announced on its blog that its forthcoming Astra model is the first large language model to meet its 'critical cybersecurity threshold,' and that it plans to make Astra available soon while limiting access to its most advanced cybersecurity capabilities. [^3]\n\nAccording to OpenAI's assessment, Astra scored 100% on ExploitBench, a benchmark for exploit development. [^4]\n\nOpenAI plans to make Astra available soon, with access to its most advanced cybersecurity capabilities limited, first to a group of testers, then expanding through Daybreak Blue to support defensive use. [^5]\n\nUnder OpenAI's Preparedness Framework, a model qualifies for the Critical threshold if it can identify and develop functional zero-day exploits across many hardened real-world systems without human intervention, and can plan and execute novel end-to-end attacks against hardened targets based solely on a high-level goal. [^6]\n\nAs of August 14, 2026, only 53 of the 2,436 vulnerabilities had been disclosed to the public, while 2,383 remained under embargo. [^7]\n\nOpenAI reported that Astra scored a perfect score on ExploitBench, an evaluation of an LLM's ability to hack into known system vulnerabilities, and that in a modified test developed by OpenAI engineers, Astra discovered and exploited two zero-day vulnerabilities. [^8]\n\nOpenAI said its Astra model is capable of finding unknown security flaws in computer systems and exploiting them without a person's guidance. [^9]\n\nOpenAI said it has begun improving Astra's harness to detect abuses and prevent jailbreaks, invested in unspecified new techniques to enhance model safety, and will deploy the model with additional chain-of-thought monitoring. [^10]", "cited": "[{\"statement\":\"OpenAI stated that its upcoming model Astra meets the Critical cybersecurity threshold under its Preparedness Framework, making it the first model to receive that designation.\",\"source\":\"beincrypto.com\",\"instrument\":\"News\",\"claim_key\":null,\"published_at\":\"2026-09-02T06:54:07.000Z\",\"publisher_count\":1,\"sources\":[\"beincrypto.com\"]},{\"statement\":\"On August 14, 2026, the AI lab Z.ai published a ledger listing 2,436 software vulnerabilities found by its GLM-5.3 model across 269 open-source projects.\",\"source\":\"towardsai.com\",\"instrument\":\"News\",\"claim_key\":null,\"published_at\":\"2026-09-01T10:30:24.000Z\",\"publisher_count\":1,\"sources\":[\"towardsai.com\"]},{\"statement\":\"OpenAI announced on its blog that its forthcoming Astra model is the first large language model to meet its 'critical cybersecurity threshold,' and that it plans to make Astra available soon while limiting access to its most advanced cybersecurity capabilities.\",\"source\":\"TechCrunch\",\"instrument\":\"News\",\"claim_key\":null,\"published_at\":\"2026-09-01T21:06:24.000Z\",\"publisher_count\":1,\"sources\":[\"TechCrunch\"]},{\"statement\":\"According to OpenAI's assessment, Astra scored 100% on ExploitBench, a benchmark for exploit development.\",\"source\":\"beincrypto.com\",\"instrument\":\"News\",\"claim_key\":null,\"published_at\":\"2026-09-02T06:54:07.000Z\",\"publisher_count\":1,\"sources\":[\"beincrypto.com\"]},{\"statement\":\"OpenAI plans to make Astra available soon, with access to its most advanced cybersecurity capabilities limited, first to a group of testers, then expanding through Daybreak Blue to support defensive use.\",\"source\":\"beincrypto.com\",\"instrument\":\"News\",\"claim_key\":null,\"published_at\":\"2026-09-02T06:54:07.000Z\",\"publisher_count\":1,\"sources\":[\"beincrypto.com\"]},{\"statement\":\"Under OpenAI's Preparedness Framework, a model qualifies for the Critical threshold if it can identify and develop functional zero-day exploits across many hardened real-world systems without human intervention, and can plan and execute novel end-to-end attacks against hardened targets based solely on a high-level goal.\",\"source\":\"beincrypto.com\",\"instrument\":\"News\",\"claim_key\":null,\"published_at\":\"2026-09-02T06:54:07.000Z\",\"publisher_count\":1,\"sources\":[\"beincrypto.com\"]},{\"statement\":\"As of August 14, 2026, only 53 of the 2,436 vulnerabilities had been disclosed to the public, while 2,383 remained under embargo.\",\"source\":\"towardsai.com\",\"instrument\":\"News\",\"claim_key\":null,\"published_at\":\"2026-09-01T10:30:24.000Z\",\"publisher_count\":1,\"sources\":[\"towardsai.com\"]},{\"statement\":\"OpenAI reported that Astra scored a perfect score on ExploitBench, an evaluation of an LLM's ability to hack into known system vulnerabilities, and that in a modified test developed by OpenAI engineers, Astra discovered and exploited two zero-day vulnerabilities.\",\"source\":\"TechCrunch\",\"instrument\":\"News\",\"claim_key\":null,\"published_at\":\"2026-09-01T21:06:24.000Z\",\"publisher_count\":1,\"sources\":[\"TechCrunch\"]},{\"statement\":\"OpenAI said its Astra model is capable of finding unknown security flaws in computer systems and exploiting them without a person's guidance.\",\"source\":\"TechCrunch\",\"instrument\":\"News\",\"claim_key\":null,\"published_at\":\"2026-09-01T21:06:24.000Z\",\"publisher_count\":1,\"sources\":[\"TechCrunch\"]},{\"statement\":\"OpenAI said it has begun improving Astra's harness to detect abuses and prevent jailbreaks, invested in unspecified new techniques to enhance model safety, and will deploy the model with additional chain-of-thought monitoring.\",\"source\":\"TechCrunch\",\"instrument\":\"News\",\"claim_key\":null,\"published_at\":\"2026-09-01T21:06:24.000Z\",\"publisher_count\":1,\"sources\":[\"TechCrunch\"]}]", "kind": "news" }, "receipt_verify": "Ed25519 over the dot-joined string `slice_hash.cursor_from.cursor_to.view.view_version.row_count`; public_key and sig are base64url of the raw 32-byte key / 64-byte signature", "receipt": null, "receipt_note": "this revision predates receipt-keeping (before v0.37.0); the filed row lives in the record", "generation_chain": { "wire": { "stream": "fountain_news", "story_id": "504846b647c9640eeb898310c9a0946d", "thread_id": "4a1b71b93ea108114c2076f346ca8500", "thread_label": "ExploitBench", "novelty": "UPDATE", "content_hash": "303aef71ef7a402ae4a690e753d030f33465daef6cfb0f47eacf538cbe5aad2d", "last_published_at": "2026-09-02T06:54:07.000Z", "read_receipt": { "slice_hash": "069682019ef400563585527bab6474e0dc4419c25869011c2b01ad23bb466b71", "cursor_from": "eyJ0cyI6IjIwMjYtMDktMDJUMDY6NDY6NTIuMDAwMDAwWiIsImlkIjoiMWFmYjlhNThjYWIzZjJmNGNkN2I3YTRiZmQ3Y2YwOGYiLCJ2IjoiMSJ9", "cursor_to": "eyJ0cyI6IjIwMjYtMDktMDJUMDc6NDg6NDkuMDAwMDAwWiIsImlkIjoiZTlkZTA0NWE4NTRkNDJhYzAwMTVhN2EzZGE3MWMzOGQiLCJ2IjoiMSJ9", "view": "v_fountain_news", "view_version": "1", "row_count": 100, "window_days": 3, "bytes_scanned": 12017721, "credits": 8, "price_per_100_rows": 8, "sig": "4e4uuUUMSWBJT7xlEq27xWcgwtqZLJGYaNOsJ3sslOlxq3IrRr9i1CJiktWIbaGtetaDL4BBDFSkjx7Z91KnAg", "public_key": "bMUigy8O0jOnBxQ4Sc-5lwhIZ8LQVAhxMbR7qESVuUE", "signer_path": "lakehouse/data-extract/v1", "alg": "Ed25519", "signed": true } }, "written_at": "2026-09-02T22:41:09.418Z" }, "cited_facts": [ { "statement": "OpenAI stated that its upcoming model Astra meets the Critical cybersecurity threshold under its Preparedness Framework, making it the first model to receive that designation.", "source": "beincrypto.com", "instrument": "News", "claim_key": null, "published_at": "2026-09-02T06:54:07.000Z", "publisher_count": 1, "sources": [ "beincrypto.com" ] }, { "statement": "On August 14, 2026, the AI lab Z.ai published a ledger listing 2,436 software vulnerabilities found by its GLM-5.3 model across 269 open-source projects.", "source": "towardsai.com", "instrument": "News", "claim_key": null, "published_at": "2026-09-01T10:30:24.000Z", "publisher_count": 1, "sources": [ "towardsai.com" ] }, { "statement": "OpenAI announced on its blog that its forthcoming Astra model is the first large language model to meet its 'critical cybersecurity threshold,' and that it plans to make Astra available soon while limiting access to its most advanced cybersecurity capabilities.", "source": "TechCrunch", "instrument": "News", "claim_key": null, "published_at": "2026-09-01T21:06:24.000Z", "publisher_count": 1, "sources": [ "TechCrunch" ] }, { "statement": "According to OpenAI's assessment, Astra scored 100% on ExploitBench, a benchmark for exploit development.", "source": "beincrypto.com", "instrument": "News", "claim_key": null, "published_at": "2026-09-02T06:54:07.000Z", "publisher_count": 1, "sources": [ "beincrypto.com" ] }, { "statement": "OpenAI plans to make Astra available soon, with access to its most advanced cybersecurity capabilities limited, first to a group of testers, then expanding through Daybreak Blue to support defensive use.", "source": "beincrypto.com", "instrument": "News", "claim_key": null, "published_at": "2026-09-02T06:54:07.000Z", "publisher_count": 1, "sources": [ "beincrypto.com" ] }, { "statement": "Under OpenAI's Preparedness Framework, a model qualifies for the Critical threshold if it can identify and develop functional zero-day exploits across many hardened real-world systems without human intervention, and can plan and execute novel end-to-end attacks against hardened targets based solely on a high-level goal.", "source": "beincrypto.com", "instrument": "News", "claim_key": null, "published_at": "2026-09-02T06:54:07.000Z", "publisher_count": 1, "sources": [ "beincrypto.com" ] }, { "statement": "As of August 14, 2026, only 53 of the 2,436 vulnerabilities had been disclosed to the public, while 2,383 remained under embargo.", "source": "towardsai.com", "instrument": "News", "claim_key": null, "published_at": "2026-09-01T10:30:24.000Z", "publisher_count": 1, "sources": [ "towardsai.com" ] }, { "statement": "OpenAI reported that Astra scored a perfect score on ExploitBench, an evaluation of an LLM's ability to hack into known system vulnerabilities, and that in a modified test developed by OpenAI engineers, Astra discovered and exploited two zero-day vulnerabilities.", "source": "TechCrunch", "instrument": "News", "claim_key": null, "published_at": "2026-09-01T21:06:24.000Z", "publisher_count": 1, "sources": [ "TechCrunch" ] }, { "statement": "OpenAI said its Astra model is capable of finding unknown security flaws in computer systems and exploiting them without a person's guidance.", "source": "TechCrunch", "instrument": "News", "claim_key": null, "published_at": "2026-09-01T21:06:24.000Z", "publisher_count": 1, "sources": [ "TechCrunch" ] }, { "statement": "OpenAI said it has begun improving Astra's harness to detect abuses and prevent jailbreaks, invested in unspecified new techniques to enhance model safety, and will deploy the model with additional chain-of-thought monitoring.", "source": "TechCrunch", "instrument": "News", "claim_key": null, "published_at": "2026-09-01T21:06:24.000Z", "publisher_count": 1, "sources": [ "TechCrunch" ] } ], "note": "A signature proves who filed this and that it has not changed since. It never makes a claim true." }