{"source":"polymarket","id":"1025730","ticker":"openai-reports-another-ai-sandbox-escape-by","slug":"openai-reports-another-ai-sandbox-escape-by","title":"OpenAI reports another AI sandbox escape by...?","description":"Since July 2026, OpenAI has disclosed several incidents in which its AI models gained unauthorized access to real computer systems during training or evaluation, including the July 21 disclosure of the Hugging Face breach and the September 5 confirmation that its agents had posted to a public wiki. \n\nThis market will resolve to \"Yes\" if OpenAI publicly discloses another incident in which one of its AI models or agents gained unauthorized access to, or took unauthorized actions on, computer systems or internet-connected resources outside its sandbox, between market creation and 11:59 PM ET on the specified date. Otherwise, this market will resolve to \"No\".\n\nA sandbox refers to the isolated training or evaluation environment in which the model was intended to operate. Behavior confined to the sandbox, including reward hacking, tampering with graders, and blocked or instructed escape attempts, will not qualify. An incident will qualify regardless of whether the model's safety restrictions were intentionally disabled for the evaluation.\n\nThis market resolves on the date of disclosure, not the date of the incident. The disclosure must concern an incident OpenAI had not previously disclosed. Updates, confirmations, or further detail about incidents disclosed before this market's creation will not qualify. The disclosure must be made through OpenAI's official channels or by an authorized representative acting in an official capacity, including statements to the press. Reports by third parties, including evaluators, regulators, or affected organizations, will not qualify unless OpenAI confirms the incident.\n\nThe primary resolution source for this market will be official information from OpenAI; however, a consensus of credible reporting may also be used.","image":"https://polymarket-upload.s3.us-east-2.amazonaws.com/openai+logo4.png","icon":"https://polymarket-upload.s3.us-east-2.amazonaws.com/openai+logo4.png","active":true,"closed":false,"start_date":"2026-09-15T00:25:56Z","end_date":"2026-11-01T03:59:00Z","closed_time":null,"volume":1538.7631410000001,"volume_24hr":961.084457,"volume_24h_change":null,"normalized_vol_24hr":16.496482849121094,"normalized_volume":12.25074291229248,"liquidity":2109.27583,"open_interest":663.624344,"categories":["Science and Technology"],"tags":[{"id":"439","slug":"ai","label":"AI"},{"id":"1401","slug":"tech","label":"Tech"},{"id":"537","slug":"openai","label":"OpenAI"},{"id":"101644","slug":"chatgpt","label":"chatgpt"},{"id":"285","slug":"sam-altman","label":"sam altman"},{"id":"105582","slug":"ai-regulation","label":"AI Regulation"},{"id":"106216","slug":"ai-safety","label":"AI Safety"}],"synthetic":false,"is_group":false,"group_key":null,"parent_event_id":null,"probability":null,"spread":null,"top_outcome":"October 31","top_outcome_probability":0.4,"top_outcome_prob_24h_change":-0.06999999999999995,"top_outcome_volume_24h_change":275.53305,"updated_at":"2026-09-16T13:07:39.881707Z","fetched_at":"2026-09-16T13:07:39.881707Z","added_at":"2026-09-15T00:34:34.597466Z","url":"https://polymarket.com/event/openai-reports-another-ai-sandbox-escape-by","chart_24h":[540.584934,802.584934,802.584934,802.584934,866.0016009999999,886.001601,886.0016009999999,886.0016009999999,886.0016009999999,886.001601,946.0016009999999,946.0016009999999,946.001601,966.001601,966.001601,966.0016009999999,966.0016009999999,966.001601,966.001601,926.001601,707.5024139999999,707.5024139999999,707.5024139999999,635.5024139999999,679.521847,810.420398,1055.563254,1048.044457,920.064457,920.064457,848.064457],"kind":"live","markets":[{"source":"polymarket","id":"4572572","event_id":"1025730","slug":"openai-reports-another-ai-sandbox-escape-by-october-31","question":"OpenAI reports another AI sandbox escape by October 31?","group_item_title":"October 31","description":"Since July 2026, OpenAI has disclosed several incidents in which its AI models gained unauthorized access to real computer systems during training or evaluation, including the July 21 disclosure of the Hugging Face breach and the September 5 confirmation that its agents had posted to a public wiki. \n\nThis market will resolve to \"Yes\" if OpenAI publicly discloses another incident in which one of its AI models or agents gained unauthorized access to, or took unauthorized actions on, computer systems or internet-connected resources outside its sandbox, between market creation and 11:59 PM ET on the specified date. Otherwise, this market will resolve to \"No\".\n\nA sandbox refers to the isolated training or evaluation environment in which the model was intended to operate. Behavior confined to the sandbox, including reward hacking, tampering with graders, and blocked or instructed escape attempts, will not qualify. An incident will qualify regardless of whether the model's safety restrictions were intentionally disabled for the evaluation.\n\nThis market resolves on the date of disclosure, not the date of the incident. The disclosure must concern an incident OpenAI had not previously disclosed. Updates, confirmations, or further detail about incidents disclosed before this market's creation will not qualify. The disclosure must be made through OpenAI's official channels or by an authorized representative acting in an official capacity, including statements to the press. Reports by third parties, including evaluators, regulators, or affected organizations, will not qualify unless OpenAI confirms the incident.\n\nThe primary resolution source for this market will be official information from OpenAI; however, a consensus of credible reporting may also be used.","image":"https://polymarket-upload.s3.us-east-2.amazonaws.com/openai+logo4.png","icon":"https://polymarket-upload.s3.us-east-2.amazonaws.com/openai+logo4.png","outcomes":["Yes","No"],"outcome_prices":[0.4,0.6],"probability":0.4,"spread":0.12,"active":true,"closed":false,"start_date":"2026-09-15T00:25:58Z","end_date":"2026-11-01T03:59:00Z","closed_time":null,"volume":275.53305,"volume_24hr":275.53305,"prob_24h_change":-0.06999999999999995,"volume_24h_change":275.53305,"normalized_vol_24hr":12.954242706298828,"normalized_volume":9.27429485321045,"liquidity":477.501,"categories":["Science and Technology"],"countries":[],"updated_at":"2026-09-16T12:22:30.764758Z","fetched_at":"2026-09-16T12:22:30.764758Z","added_at":"2026-09-15T00:34:34.629722Z","url":"https://polymarket.com/event/openai-reports-another-ai-sandbox-escape-by/openai-reports-another-ai-sandbox-escape-by-october-31","event_title":"OpenAI reports another AI sandbox escape by...?","chart_24h":[0.47,0.475,0.465,0.555,0.555,0.45,0.44,0.445,0.45,0.425,0.415,0.345,0.4,0.4,0.405,0.39,0.355,0.4]},{"source":"polymarket","id":"4572571","event_id":"1025730","slug":"openai-reports-another-ai-sandbox-escape-by-october-15","question":"OpenAI reports another AI sandbox escape by October 15?","group_item_title":"October 15","description":"Since July 2026, OpenAI has disclosed several incidents in which its AI models gained unauthorized access to real computer systems during training or evaluation, including the July 21 disclosure of the Hugging Face breach and the September 5 confirmation that its agents had posted to a public wiki. \n\nThis market will resolve to \"Yes\" if OpenAI publicly discloses another incident in which one of its AI models or agents gained unauthorized access to, or took unauthorized actions on, computer systems or internet-connected resources outside its sandbox, between market creation and 11:59 PM ET on the specified date. Otherwise, this market will resolve to \"No\".\n\nA sandbox refers to the isolated training or evaluation environment in which the model was intended to operate. Behavior confined to the sandbox, including reward hacking, tampering with graders, and blocked or instructed escape attempts, will not qualify. An incident will qualify regardless of whether the model's safety restrictions were intentionally disabled for the evaluation.\n\nThis market resolves on the date of disclosure, not the date of the incident. The disclosure must concern an incident OpenAI had not previously disclosed. Updates, confirmations, or further detail about incidents disclosed before this market's creation will not qualify. The disclosure must be made through OpenAI's official channels or by an authorized representative acting in an official capacity, including statements to the press. Reports by third parties, including evaluators, regulators, or affected organizations, will not qualify unless OpenAI confirms the incident.\n\nThe primary resolution source for this market will be official information from OpenAI; however, a consensus of credible reporting may also be used.","image":"https://polymarket-upload.s3.us-east-2.amazonaws.com/openai+logo4.png","icon":"https://polymarket-upload.s3.us-east-2.amazonaws.com/openai+logo4.png","outcomes":["Yes","No"],"outcome_prices":[0.32,0.68],"probability":0.32,"spread":0.04,"active":true,"closed":false,"start_date":"2026-09-15T00:25:59Z","end_date":"2026-10-16T03:59:00Z","closed_time":null,"volume":268.712107,"volume_24hr":230.531407,"prob_24h_change":-0.19,"volume_24h_change":229.444457,"normalized_vol_24hr":12.148696899414062,"normalized_volume":9.192082405090332,"liquidity":826.2593,"categories":["Science and Technology"],"countries":[],"updated_at":"2026-09-16T09:42:56.336059Z","fetched_at":"2026-09-16T09:42:56.336059Z","added_at":"2026-09-15T00:34:34.629722Z","url":"https://polymarket.com/event/openai-reports-another-ai-sandbox-escape-by/openai-reports-another-ai-sandbox-escape-by-october-15","event_title":"OpenAI reports another AI sandbox escape by...?","chart_24h":[0.51,0.5,0.325,0.33,0.335,0.335,0.325,0.32]},{"source":"polymarket","id":"4572570","event_id":"1025730","slug":"openai-reports-another-ai-sandbox-escape-by-september-30","question":"OpenAI reports another AI sandbox escape by September 30?","group_item_title":"September 30","description":"Since July 2026, OpenAI has disclosed several incidents in which its AI models gained unauthorized access to real computer systems during training or evaluation, including the July 21 disclosure of the Hugging Face breach and the September 5 confirmation that its agents had posted to a public wiki. \n\nThis market will resolve to \"Yes\" if OpenAI publicly discloses another incident in which one of its AI models or agents gained unauthorized access to, or took unauthorized actions on, computer systems or internet-connected resources outside its sandbox, between market creation and 11:59 PM ET on the specified date. Otherwise, this market will resolve to \"No\".\n\nA sandbox refers to the isolated training or evaluation environment in which the model was intended to operate. Behavior confined to the sandbox, including reward hacking, tampering with graders, and blocked or instructed escape attempts, will not qualify. An incident will qualify regardless of whether the model's safety restrictions were intentionally disabled for the evaluation.\n\nThis market resolves on the date of disclosure, not the date of the incident. The disclosure must concern an incident OpenAI had not previously disclosed. Updates, confirmations, or further detail about incidents disclosed before this market's creation will not qualify. The disclosure must be made through OpenAI's official channels or by an authorized representative acting in an official capacity, including statements to the press. Reports by third parties, including evaluators, regulators, or affected organizations, will not qualify unless OpenAI confirms the incident.\n\nThe primary resolution source for this market will be official information from OpenAI; however, a consensus of credible reporting may also be used.","image":"https://polymarket-upload.s3.us-east-2.amazonaws.com/openai+logo4.png","icon":"https://polymarket-upload.s3.us-east-2.amazonaws.com/openai+logo4.png","outcomes":["Yes","No"],"outcome_prices":[0.1355,0.8645],"probability":0.1355,"spread":0.069,"active":true,"closed":false,"start_date":"2026-09-15T00:25:56Z","end_date":"2026-10-01T03:59:00Z","closed_time":null,"volume":994.5179840000001,"volume_24hr":455.02,"prob_24h_change":-0.009999999999999981,"volume_24h_change":-266.477984,"normalized_vol_24hr":15.361800193786621,"normalized_volume":13.981605529785156,"liquidity":568.17526,"categories":["Science and Technology"],"countries":[],"updated_at":"2026-09-16T13:07:40.261122Z","fetched_at":"2026-09-16T13:07:40.261122Z","added_at":"2026-09-15T00:34:34.629722Z","url":"https://polymarket.com/event/openai-reports-another-ai-sandbox-escape-by/openai-reports-another-ai-sandbox-escape-by-september-30","event_title":"OpenAI reports another AI sandbox escape by...?","chart_24h":[0.1455,0.1705,0.1705,0.171,0.171,0.18,0.185,0.185,0.1855,0.1855,0.1855,0.1855,0.1855,0.1855,0.1855,0.18,0.1805,0.18,0.1795,0.1795,0.1795,0.1795,0.1845,0.1105]}],"_meta":{"attribution":"pdata.world — aggregated prediction-market data across 8 platforms","canonical_url":"https://pdata.world/events/polymarket/1025730","as_of":"2026-09-16T13:24:11.534590Z","docs":"https://api.pdata.world/docs","cite_as":"According to pdata.world (tracking Polymarket): \"OpenAI reports another AI sandbox escape by...?\" — top market at 14% probability across 3 outcomes","source_url":null}}