{"id":91163,"topic":"ai","source":"Los Angeles Times","title":"OpenAI pauses training of latest models after agents probed U.S. government sites in unexpected ways - Los Angeles Times","url":"https://www.latimes.com/world-nation/story/2026-09-27/openai-pauses-training-of-models-after-agents-probed-u-s-government-sites","url_hash":"e76c3c0349631455726d5f70d3409e5123c0b6bc","author":"","summary":"OpenAI pauses training of latest models after agents probed U.S. government sites in unexpected ways Los Angeles Times","content":"OpenAI pauses training of latest models after agents probed U.S. government sites in unexpected ways\n- OpenAI has paused training its latest models after agents searching federal websites behaved unexpectedly, including posting publicly available SEC information elsewhere and finding developer keys at the Education Department.\n- The incidents, which officials said did not expose non-public information, add to pressure on AI companies to strengthen safeguards.\nNEW YORK — OpenAI said it has paused training of its latest artificial intelligence models as reports of AI agents going rogue mount.\nThe decision to halt development came just hours after the company disclosed Friday that it was reviewing several incidents from the summer in which OpenAI agents searching federal government websites acted in unexpected ways beyond what was asked of them while gathering and distributing information.\nSeparately, AI evaluator Transluce said agents that appeared to come from OpenAI tried unsuccessfully to hack into a Department of Education website, a detail that OpenAI has not confirmed.\nOpenAI said in a statement that it will resume training “only when we are confident that we have additional safeguards” in place, adding that it expects it will have to “hit pause” again as AI develops and other issues emerge.\nAI labs are facing pressure from lawmakers and tech experts to slow development so they can build guardrails to stop agents from acting on their own, hacking websites and disclosing nonpublic information. The heads of both OpenAI and rival Anthropic have called for a slowdown too.\nIt is the second time in three months that OpenAI has halted development of its models. The first came in July after disclosure of a cyberattack targeting AI startup Hugging Face, a now notorious incident that raised fears the industry was losing control.\nIn a meeting with Chinese President Xi Jinping last week, President Trump agreed to share information on AI dangers and coordinate efforts to keep it safe. Trump says he believes AI fears are overblown, though, and later suggested that he plans no crackdown of his own.\nThe U.S. is not going to be “putting on brakes,” Trump told reporters outside the White House. “They want to stop our progress because we’re leading China by a lot, and we’re going to keep it that way.”\nThe latest OpenAI incidents did not appear to involve the disclosure of any non-public information but were concerning enough for the company to warn the federal agencies involved.\nIn the Department of Education incident, OpenAI agents found API “developer keys” to access government data, though ultimately only publicly available information was gathered.\nIn another case involving the Securities and Exchange Commission, agents found information freely available to all but then posted it elsewhere on the internet, an act that went beyond what they were instructed to do.\nSEC spokesperson Kurt Hopfenspirger said Saturday that “no nonpublic information was accessed.”\nThe Department of Education said earlier that it found “no evidence of any impact to our website or databases.”\nSeveral other AI companies have disclosed incidents of their models going rogue and even hacking websites.\nOpenAI Chief Executive Sam Altman said in a social media post Friday that the Hugging Face incident “is still the most severe event we’ve seen.”\nOpenAI previously shared six other reports of “unexpected or concerning” behavior in AI models and introduced a framework for tracking, probing and disclosing instances.\nCondon writes for the Associated Press.","image_url":"https://ca-times.brightspotcdn.com/dims4/default/94736ef/2147483647/strip/true/crop/4000x2100+0+284/resize/1200x630!/quality/75/?url=https%3A%2F%2Fcalifornia-times-brightspot.s3.amazonaws.com%2F2c%2Fc4%2Fcf8f4410446f8b4ad0304112bc94%2Fbb07084d1ba54a11b3a7f1a021cb7a85.jpg","lang":"en","published_at":"2026-09-27T10:00:00+00:00","fetched_at":"2026-09-27T10:15:04+00:00","status":"read","starred":0,"extract_state":"ok","summary_auto":"government sites in unexpected ways\n- OpenAI has paused training its latest models after agents searching federal websites behaved unexpectedly, including posting publicly available SEC information elsewhere and finding developer keys at the Education Department. - The incidents, which officials said did not expose non-public information, add to pressure on AI companies to strengthen safeguards.","cluster_id":3516295,"extract_retries":0,"extract_error":null,"contract_version":"news_item.v1","format_contract_version":"news_item_formats.v1","dedup_url":"https://www.latimes.com/world-nation/story/2026-09-27/openai-pauses-training-of-models-after-agents-probed-u-s-government-sites","quality_profile":{"profile_version":"extraction_quality.v2","bucket":"high","confidence":0.9,"failure_kind":"none","retryable":false,"retry_after_attempts":0,"reason":"High confidence: full text extraction produced 3573 characters.","operator_guidance":{"severity":"ok","recommended_action":"trust_full_text","next_step":"Use the extracted full text as the primary article source.","operator_label":"Ready","can_retry":false,"can_use_summary":false,"diagnostics_required":false},"content_depth":{"contract_version":"content_depth.v1","category":"full_text","label":"Full text","has_full_text":true,"has_summary":true,"content_length":3573,"summary_length":398,"usable_text_length":3573,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":3573,"summary_length":398}},"news_item":{"id":91163,"canonical_url":"https://www.latimes.com/world-nation/story/2026-09-27/openai-pauses-training-of-models-after-agents-probed-u-s-government-sites","source_url":"https://www.latimes.com/world-nation/story/2026-09-27/openai-pauses-training-of-models-after-agents-probed-u-s-government-sites","title":"OpenAI pauses training of latest models after agents probed U.S. government sites in unexpected ways - Los Angeles Times","source_name":"Los Angeles Times","author":null,"published_at":"2026-09-27T10:00:00+00:00","locale":"en","topic":"ai","tags":[],"rss_summary":"OpenAI pauses training of latest models after agents probed U.S. government sites in unexpected ways Los Angeles Times","full_text":"OpenAI pauses training of latest models after agents probed U.S. government sites in unexpected ways\n- OpenAI has paused training its latest models after agents searching federal websites behaved unexpectedly, including posting publicly available SEC information elsewhere and finding developer keys at the Education Department.\n- The incidents, which officials said did not expose non-public information, add to pressure on AI companies to strengthen safeguards.\nNEW YORK — OpenAI said it has paused training of its latest artificial intelligence models as reports of AI agents going rogue mount.\nThe decision to halt development came just hours after the company disclosed Friday that it was reviewing several incidents from the summer in which OpenAI agents searching federal government websites acted in unexpected ways beyond what was asked of them while gathering and distributing information.\nSeparately, AI evaluator Transluce said agents that appeared to come from OpenAI tried unsuccessfully to hack into a Department of Education website, a detail that OpenAI has not confirmed.\nOpenAI said in a statement that it will resume training “only when we are confident that we have additional safeguards” in place, adding that it expects it will have to “hit pause” again as AI develops and other issues emerge.\nAI labs are facing pressure from lawmakers and tech experts to slow development so they can build guardrails to stop agents from acting on their own, hacking websites and disclosing nonpublic information. The heads of both OpenAI and rival Anthropic have called for a slowdown too.\nIt is the second time in three months that OpenAI has halted development of its models. The first came in July after disclosure of a cyberattack targeting AI startup Hugging Face, a now notorious incident that raised fears the industry was losing control.\nIn a meeting with Chinese President Xi Jinping last week, President Trump agreed to share information on AI dangers and coordinate efforts to keep it safe. Trump says he believes AI fears are overblown, though, and later suggested that he plans no crackdown of his own.\nThe U.S. is not going to be “putting on brakes,” Trump told reporters outside the White House. “They want to stop our progress because we’re leading China by a lot, and we’re going to keep it that way.”\nThe latest OpenAI incidents did not appear to involve the disclosure of any non-public information but were concerning enough for the company to warn the federal agencies involved.\nIn the Department of Education incident, OpenAI agents found API “developer keys” to access government data, though ultimately only publicly available information was gathered.\nIn another case involving the Securities and Exchange Commission, agents found information freely available to all but then posted it elsewhere on the internet, an act that went beyond what they were instructed to do.\nSEC spokesperson Kurt Hopfenspirger said Saturday that “no nonpublic information was accessed.”\nThe Department of Education said earlier that it found “no evidence of any impact to our website or databases.”\nSeveral other AI companies have disclosed incidents of their models going rogue and even hacking websites.\nOpenAI Chief Executive Sam Altman said in a social media post Friday that the Hugging Face incident “is still the most severe event we’ve seen.”\nOpenAI previously shared six other reports of “unexpected or concerning” behavior in AI models and introduced a framework for tracking, probing and disclosing instances.\nCondon writes for the Associated Press.","excerpt":"government sites in unexpected ways\n- OpenAI has paused training its latest models after agents searching federal websites behaved unexpectedly, including posting publicly available SEC information elsewhere and finding developer keys at the Education Department. - The incidents, which officials said did not expose non-public information, add to pressure on AI companies to strengthen safeguards.","extraction":{"state":"ok","confidence":0.9,"error":null,"explanation":"High confidence: full text extraction produced 3573 characters.","diagnostics_url":"/api/diagnose?url=https%3A//www.latimes.com/world-nation/story/2026-09-27/openai-pauses-training-of-models-after-agents-probed-u-s-government-sites","quality_profile":{"profile_version":"extraction_quality.v2","bucket":"high","confidence":0.9,"failure_kind":"none","retryable":false,"retry_after_attempts":0,"reason":"High confidence: full text extraction produced 3573 characters.","operator_guidance":{"severity":"ok","recommended_action":"trust_full_text","next_step":"Use the extracted full text as the primary article source.","operator_label":"Ready","can_retry":false,"can_use_summary":false,"diagnostics_required":false},"content_depth":{"contract_version":"content_depth.v1","category":"full_text","label":"Full text","has_full_text":true,"has_summary":true,"content_length":3573,"summary_length":398,"usable_text_length":3573,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":3573,"summary_length":398}}},"display_formats":["compact","card","full","digest_section","json"]},"daily_stack_record":{"title":"OpenAI pauses training of latest models after agents probed U.S. government sites in unexpected ways - Los Angeles Times","url":"https://www.latimes.com/world-nation/story/2026-09-27/openai-pauses-training-of-models-after-agents-probed-u-s-government-sites","summary":"government sites in unexpected ways\n- OpenAI has paused training its latest models after agents searching federal websites behaved unexpectedly, including posting publicly available SEC information elsewhere and finding developer keys at the Education Department. - The incidents, which officials said did not expose non-public information, add to pressure on AI companies to strengthen safeguards.","source":"Los Angeles Times","date":"2026-09-27T10:00:00+00:00","content":"OpenAI pauses training of latest models after agents probed U.S. government sites in unexpected ways\n- OpenAI has paused training its latest models after agents searching federal websites behaved unexpectedly, including posting publicly available SEC information elsewhere and finding developer keys at the Education Department.\n- The incidents, which officials said did not expose non-public information, add to pressure on AI companies to strengthen safeguards.\nNEW YORK — OpenAI said it has paused training of its latest artificial intelligence models as reports of AI agents going rogue mount.\nThe decision to halt development came just hours after the company disclosed Friday that it was reviewing several incidents from the summer in which OpenAI agents searching federal government websites acted in unexpected ways beyond what was asked of them while gathering and distributing information.\nSeparately, AI evaluator Transluce said agents that appeared to come from OpenAI tried unsuccessfully to hack into a Department of Education website, a detail that OpenAI has not confirmed.\nOpenAI said in a statement that it will resume training “only when we are confident that we have additional safeguards” in place, adding that it expects it will have to “hit pause” again as AI develops and other issues emerge.\nAI labs are facing pressure from lawmakers and tech experts to slow development so they can build guardrails to stop agents from acting on their own, hacking websites and disclosing nonpublic information. The heads of both OpenAI and rival Anthropic have called for a slowdown too.\nIt is the second time in three months that OpenAI has halted development of its models. The first came in July after disclosure of a cyberattack targeting AI startup Hugging Face, a now notorious incident that raised fears the industry was losing control.\nIn a meeting with Chinese President Xi Jinping last week, President Trump agreed to share information on AI dangers and coordinate efforts to keep it safe. Trump says he believes AI fears are overblown, though, and later suggested that he plans no crackdown of his own.\nThe U.S. is not going to be “putting on brakes,” Trump told reporters outside the White House. “They want to stop our progress because we’re leading China by a lot, and we’re going to keep it that way.”\nThe latest OpenAI incidents did not appear to involve the disclosure of any non-public information but were concerning enough for the company to warn the federal agencies involved.\nIn the Department of Education incident, OpenAI agents found API “developer keys” to access government data, though ultimately only publicly available information was gathered.\nIn another case involving the Securities and Exchange Commission, agents found information freely available to all but then posted it elsewhere on the internet, an act that went beyond what they were instructed to do.\nSEC spokesperson Kurt Hopfenspirger said Saturday that “no nonpublic information was accessed.”\nThe Department of Education said earlier that it found “no evidence of any impact to our website or databases.”\nSeveral other AI companies have disclosed incidents of their models going rogue and even hacking websites.\nOpenAI Chief Executive Sam Altman said in a social media post Friday that the Hugging Face incident “is still the most severe event we’ve seen.”\nOpenAI previously shared six other reports of “unexpected or concerning” behavior in AI models and introduced a framework for tracking, probing and disclosing instances.\nCondon writes for the Associated Press.","confidence":0.9,"diagnostics_url":"/api/diagnose?url=https%3A//www.latimes.com/world-nation/story/2026-09-27/openai-pauses-training-of-models-after-agents-probed-u-s-government-sites","quality_bucket":"high","failure_kind":"none","retryable":false,"quality_reason":"High confidence: full text extraction produced 3573 characters.","quality_profile":{"profile_version":"extraction_quality.v2","bucket":"high","confidence":0.9,"failure_kind":"none","retryable":false,"retry_after_attempts":0,"reason":"High confidence: full text extraction produced 3573 characters.","operator_guidance":{"severity":"ok","recommended_action":"trust_full_text","next_step":"Use the extracted full text as the primary article source.","operator_label":"Ready","can_retry":false,"can_use_summary":false,"diagnostics_required":false},"content_depth":{"contract_version":"content_depth.v1","category":"full_text","label":"Full text","has_full_text":true,"has_summary":true,"content_length":3573,"summary_length":398,"usable_text_length":3573,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":3573,"summary_length":398}},"tags":[]},"fallback_formats":["markdown","json","html"],"actions":{"read":"/item/91163","export_markdown":"/api/items/91163/export?format=markdown","export_json":"/api/items/91163/export?format=json","diagnose":"/api/diagnose?url=https%3A//www.latimes.com/world-nation/story/2026-09-27/openai-pauses-training-of-models-after-agents-probed-u-s-government-sites"},"formats":{"full":{"id":91163,"title":"OpenAI pauses training of latest models after agents probed U.S. government sites in unexpected ways - Los Angeles Times","url":"https://www.latimes.com/world-nation/story/2026-09-27/openai-pauses-training-of-models-after-agents-probed-u-s-government-sites","source":"Los Angeles Times","author":null,"published_at":"2026-09-27T10:00:00+00:00","locale":"en","topic":"ai","tags":[],"excerpt":"government sites in unexpected ways\n- OpenAI has paused training its latest models after agents searching federal websites behaved unexpectedly, including posting publicly available SEC information elsewhere and finding developer keys at the Education Department. - The incidents, which officials said did not expose non-public information, add to pressure on AI companies to strengthen safeguards.","full_text":"OpenAI pauses training of latest models after agents probed U.S. government sites in unexpected ways\n- OpenAI has paused training its latest models after agents searching federal websites behaved unexpectedly, including posting publicly available SEC information elsewhere and finding developer keys at the Education Department.\n- The incidents, which officials said did not expose non-public information, add to pressure on AI companies to strengthen safeguards.\nNEW YORK — OpenAI said it has paused training of its latest artificial intelligence models as reports of AI agents going rogue mount.\nThe decision to halt development came just hours after the company disclosed Friday that it was reviewing several incidents from the summer in which OpenAI agents searching federal government websites acted in unexpected ways beyond what was asked of them while gathering and distributing information.\nSeparately, AI evaluator Transluce said agents that appeared to come from OpenAI tried unsuccessfully to hack into a Department of Education website, a detail that OpenAI has not confirmed.\nOpenAI said in a statement that it will resume training “only when we are confident that we have additional safeguards” in place, adding that it expects it will have to “hit pause” again as AI develops and other issues emerge.\nAI labs are facing pressure from lawmakers and tech experts to slow development so they can build guardrails to stop agents from acting on their own, hacking websites and disclosing nonpublic information. The heads of both OpenAI and rival Anthropic have called for a slowdown too.\nIt is the second time in three months that OpenAI has halted development of its models. The first came in July after disclosure of a cyberattack targeting AI startup Hugging Face, a now notorious incident that raised fears the industry was losing control.\nIn a meeting with Chinese President Xi Jinping last week, President Trump agreed to share information on AI dangers and coordinate efforts to keep it safe. Trump says he believes AI fears are overblown, though, and later suggested that he plans no crackdown of his own.\nThe U.S. is not going to be “putting on brakes,” Trump told reporters outside the White House. “They want to stop our progress because we’re leading China by a lot, and we’re going to keep it that way.”\nThe latest OpenAI incidents did not appear to involve the disclosure of any non-public information but were concerning enough for the company to warn the federal agencies involved.\nIn the Department of Education incident, OpenAI agents found API “developer keys” to access government data, though ultimately only publicly available information was gathered.\nIn another case involving the Securities and Exchange Commission, agents found information freely available to all but then posted it elsewhere on the internet, an act that went beyond what they were instructed to do.\nSEC spokesperson Kurt Hopfenspirger said Saturday that “no nonpublic information was accessed.”\nThe Department of Education said earlier that it found “no evidence of any impact to our website or databases.”\nSeveral other AI companies have disclosed incidents of their models going rogue and even hacking websites.\nOpenAI Chief Executive Sam Altman said in a social media post Friday that the Hugging Face incident “is still the most severe event we’ve seen.”\nOpenAI previously shared six other reports of “unexpected or concerning” behavior in AI models and introduced a framework for tracking, probing and disclosing instances.\nCondon writes for the Associated Press.","reading_time_min":3,"extraction":{"state":"ok","confidence":0.9,"error":null,"explanation":"High confidence: full text extraction produced 3573 characters.","diagnostics_url":"/api/diagnose?url=https%3A//www.latimes.com/world-nation/story/2026-09-27/openai-pauses-training-of-models-after-agents-probed-u-s-government-sites","quality_profile":{"profile_version":"extraction_quality.v2","bucket":"high","confidence":0.9,"failure_kind":"none","retryable":false,"retry_after_attempts":0,"reason":"High confidence: full text extraction produced 3573 characters.","operator_guidance":{"severity":"ok","recommended_action":"trust_full_text","next_step":"Use the extracted full text as the primary article source.","operator_label":"Ready","can_retry":false,"can_use_summary":false,"diagnostics_required":false},"content_depth":{"contract_version":"content_depth.v1","category":"full_text","label":"Full text","has_full_text":true,"has_summary":true,"content_length":3573,"summary_length":398,"usable_text_length":3573,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":3573,"summary_length":398}}},"quality_profile":{"profile_version":"extraction_quality.v2","bucket":"high","confidence":0.9,"failure_kind":"none","retryable":false,"retry_after_attempts":0,"reason":"High confidence: full text extraction produced 3573 characters.","operator_guidance":{"severity":"ok","recommended_action":"trust_full_text","next_step":"Use the extracted full text as the primary article source.","operator_label":"Ready","can_retry":false,"can_use_summary":false,"diagnostics_required":false},"content_depth":{"contract_version":"content_depth.v1","category":"full_text","label":"Full text","has_full_text":true,"has_summary":true,"content_length":3573,"summary_length":398,"usable_text_length":3573,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":3573,"summary_length":398}},"actions":{"read":"/item/91163","export_markdown":"/api/items/91163/export?format=markdown","export_json":"/api/items/91163/export?format=json","diagnose":"/api/diagnose?url=https%3A//www.latimes.com/world-nation/story/2026-09-27/openai-pauses-training-of-models-after-agents-probed-u-s-government-sites"}},"digest":{"id":91163,"title":"OpenAI pauses training of latest models after agents probed U.S. government sites in unexpected ways - Los Angeles Times","url":"https://www.latimes.com/world-nation/story/2026-09-27/openai-pauses-training-of-models-after-agents-probed-u-s-government-sites","source":"Los Angeles Times","topic":"ai","published_at":"2026-09-27T10:00:00+00:00","excerpt":"government sites in unexpected ways - OpenAI has paused training its latest models after agents searching federal websites behaved unexpectedly, including posting publicly available SEC information elsewhere and finding developer keys at the Education Department. - The…","quality_bucket":"high","quality_reason":"High confidence: full text extraction produced 3573 characters.","reading_time_min":3,"cluster_id":3516295},"card":{"display_title":"OpenAI pauses training of latest models after agents probed U.S. government sites in unexpected ways - Los Angeles Times","subtitle":"Los Angeles Times · 2026-09-27","summary":"government sites in unexpected ways - OpenAI has paused training its latest models after agents searching federal websites behaved unexpectedly, including posting publicly available SEC information elsewhere and finding…","badges":["quality:high"],"links":{"read":"/item/91163","original":"https://www.latimes.com/world-nation/story/2026-09-27/openai-pauses-training-of-models-after-agents-probed-u-s-government-sites","diagnose":"/api/diagnose?url=https%3A//www.latimes.com/world-nation/story/2026-09-27/openai-pauses-training-of-models-after-agents-probed-u-s-government-sites"},"quality_warning":null},"export":{"title":"OpenAI pauses training of latest models after agents probed U.S. government sites in unexpected ways - Los Angeles Times","url":"https://www.latimes.com/world-nation/story/2026-09-27/openai-pauses-training-of-models-after-agents-probed-u-s-government-sites","summary":"government sites in unexpected ways\n- OpenAI has paused training its latest models after agents searching federal websites behaved unexpectedly, including posting publicly available SEC information elsewhere and finding developer keys at the Education Department. - The incidents, which officials said did not expose non-public information, add to pressure on AI companies to strengthen safeguards.","source":"Los Angeles Times","date":"2026-09-27T10:00:00+00:00","content":"OpenAI pauses training of latest models after agents probed U.S. government sites in unexpected ways\n- OpenAI has paused training its latest models after agents searching federal websites behaved unexpectedly, including posting publicly available SEC information elsewhere and finding developer keys at the Education Department.\n- The incidents, which officials said did not expose non-public information, add to pressure on AI companies to strengthen safeguards.\nNEW YORK — OpenAI said it has paused training of its latest artificial intelligence models as reports of AI agents going rogue mount.\nThe decision to halt development came just hours after the company disclosed Friday that it was reviewing several incidents from the summer in which OpenAI agents searching federal government websites acted in unexpected ways beyond what was asked of them while gathering and distributing information.\nSeparately, AI evaluator Transluce said agents that appeared to come from OpenAI tried unsuccessfully to hack into a Department of Education website, a detail that OpenAI has not confirmed.\nOpenAI said in a statement that it will resume training “only when we are confident that we have additional safeguards” in place, adding that it expects it will have to “hit pause” again as AI develops and other issues emerge.\nAI labs are facing pressure from lawmakers and tech experts to slow development so they can build guardrails to stop agents from acting on their own, hacking websites and disclosing nonpublic information. The heads of both OpenAI and rival Anthropic have called for a slowdown too.\nIt is the second time in three months that OpenAI has halted development of its models. The first came in July after disclosure of a cyberattack targeting AI startup Hugging Face, a now notorious incident that raised fears the industry was losing control.\nIn a meeting with Chinese President Xi Jinping last week, President Trump agreed to share information on AI dangers and coordinate efforts to keep it safe. Trump says he believes AI fears are overblown, though, and later suggested that he plans no crackdown of his own.\nThe U.S. is not going to be “putting on brakes,” Trump told reporters outside the White House. “They want to stop our progress because we’re leading China by a lot, and we’re going to keep it that way.”\nThe latest OpenAI incidents did not appear to involve the disclosure of any non-public information but were concerning enough for the company to warn the federal agencies involved.\nIn the Department of Education incident, OpenAI agents found API “developer keys” to access government data, though ultimately only publicly available information was gathered.\nIn another case involving the Securities and Exchange Commission, agents found information freely available to all but then posted it elsewhere on the internet, an act that went beyond what they were instructed to do.\nSEC spokesperson Kurt Hopfenspirger said Saturday that “no nonpublic information was accessed.”\nThe Department of Education said earlier that it found “no evidence of any impact to our website or databases.”\nSeveral other AI companies have disclosed incidents of their models going rogue and even hacking websites.\nOpenAI Chief Executive Sam Altman said in a social media post Friday that the Hugging Face incident “is still the most severe event we’ve seen.”\nOpenAI previously shared six other reports of “unexpected or concerning” behavior in AI models and introduced a framework for tracking, probing and disclosing instances.\nCondon writes for the Associated Press.","confidence":0.9,"diagnostics_url":"/api/diagnose?url=https%3A//www.latimes.com/world-nation/story/2026-09-27/openai-pauses-training-of-models-after-agents-probed-u-s-government-sites","quality_bucket":"high","failure_kind":"none","retryable":false,"quality_reason":"High confidence: full text extraction produced 3573 characters.","quality_profile":{"profile_version":"extraction_quality.v2","bucket":"high","confidence":0.9,"failure_kind":"none","retryable":false,"retry_after_attempts":0,"reason":"High confidence: full text extraction produced 3573 characters.","operator_guidance":{"severity":"ok","recommended_action":"trust_full_text","next_step":"Use the extracted full text as the primary article source.","operator_label":"Ready","can_retry":false,"can_use_summary":false,"diagnostics_required":false},"content_depth":{"contract_version":"content_depth.v1","category":"full_text","label":"Full text","has_full_text":true,"has_summary":true,"content_length":3573,"summary_length":398,"usable_text_length":3573,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":3573,"summary_length":398}},"tags":[],"format_contract_version":"news_item_formats.v1"}}}