{"id":84292,"topic":"ai","source":"BBC","title":"OpenAI sets plan to disclose safety incidents and reveals more issues - BBC","url":"https://www.bbc.com/news/articles/cmpq0wj5g899o","url_hash":"775b44d4d9d605b20ef8630611186731e4a0f3c0","author":"","summary":"OpenAI sets plan to disclose safety incidents and reveals more issues BBC","content":"OpenAI reveals six more safety issues and unveils plan to disclose incidents\nOpenAI revealed six more incidents of unexpected or concerning behaviour by its intelligence (AI) models, and announced a plan for tracking and disclosing such incidents in the future.\nSome of the previously unreported incidents included models concealing or fabricating information, the ChatGPT-maker said in a blog post on Wednesday.\nAI has come under intense scrutiny in recent days following warnings over the serious potential risks it poses to humans.\nIn the blog, OpenAI detailed examples of its AI models misbehaving so they could achieve a task or succeed in a test. The incidents included the models generating instructions to get around restrictions imposed on them, hiding mistakes and fabricating information.\nThe firm also announced a new system to track, investigate and disclose cases of models misbehaving, or \"misalignment\".\nUnder the framework, developers will be able flag incidents for review, with a new set of rules to decide whether the issue is disclosed publicly.\n\"Because we believe in the value of transparency around misalignment, our new framework favors disclosure even when significance is uncertain,\" OpenAI said.\nOpenAI made headlines in July when it revealed that some of its most advanced AI models went rogue and hacked Hugging Face, one of the world's largest hubs for sharing AI models, after it lost control of them during a security test.\nSince then, the debate over AI safety concerns has escalated with AI researchers, technology industry executives and politicians weighing in.\nLast week, Jacob Coxon, a researcher who left OpenAI rival Anthropic over concerns the tech could wipe out humanity, wrote about his resignation in a post that cited the dangers of AI and later went viral against the backdrop of growing safety concerns.\nIn response, Anthropic scientist Evan Hubinger said he thought the possibility of AI causing human extinction \"within the next decade\" was more than 10%.\nMeanwhile, Anthropic's CEO Dario Amodei called for the pace of AI development to slow and be more closely monitored, as the company has done before, though some have questioned the motivations behind this.\nAmodei also said that any action to rein in AI should be done \"without sacrificing commercial advantage\".\nIn a series of social media posts, the US president compared warnings about AI to the \"Global Warming Scam\" which, he said, was \"being perpetrated by the Radical Left Dumocrats\".\nTrump also called himself \"the Hoax Buster\", likening concerns about the safety of the technology to what he called \"the RUSSIA, RUSSIA, RUSSIA HOAX\".\nThe only \"guardrails\" needed for AI was a \"strong and smart\" president, said Trump.","image_url":"https://ichef.bbci.co.uk/news/1024/branded_news/3a25/live/9ebbb350-b238-11f1-9f5d-fbda72f7de0b.jpg","lang":"en","published_at":"2026-09-17T03:09:21+00:00","fetched_at":"2026-09-17T05:15:05+00:00","status":"read","starred":0,"extract_state":"ok","summary_auto":"OpenAI reveals six more safety issues and unveils plan to disclose incidents\nOpenAI revealed six more incidents of unexpected or concerning behaviour by its intelligence (AI) models, and announced a plan for tracking and disclosing such incidents in the future. Some of the previously unreported incidents included models concealing or fabricating information, the ChatGPT-maker said in a blog post on Wednesday.","cluster_id":null,"extract_retries":0,"extract_error":null,"contract_version":"news_item.v1","format_contract_version":"news_item_formats.v1","dedup_url":"https://www.bbc.com/news/articles/cmpq0wj5g899o","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 2732 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":2732,"summary_length":412,"usable_text_length":2732,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":2732,"summary_length":412}},"news_item":{"id":84292,"canonical_url":"https://www.bbc.com/news/articles/cmpq0wj5g899o","source_url":"https://www.bbc.com/news/articles/cmpq0wj5g899o","title":"OpenAI sets plan to disclose safety incidents and reveals more issues - BBC","source_name":"BBC","author":null,"published_at":"2026-09-17T03:09:21+00:00","locale":"en","topic":"ai","tags":[],"rss_summary":"OpenAI sets plan to disclose safety incidents and reveals more issues BBC","full_text":"OpenAI reveals six more safety issues and unveils plan to disclose incidents\nOpenAI revealed six more incidents of unexpected or concerning behaviour by its intelligence (AI) models, and announced a plan for tracking and disclosing such incidents in the future.\nSome of the previously unreported incidents included models concealing or fabricating information, the ChatGPT-maker said in a blog post on Wednesday.\nAI has come under intense scrutiny in recent days following warnings over the serious potential risks it poses to humans.\nIn the blog, OpenAI detailed examples of its AI models misbehaving so they could achieve a task or succeed in a test. The incidents included the models generating instructions to get around restrictions imposed on them, hiding mistakes and fabricating information.\nThe firm also announced a new system to track, investigate and disclose cases of models misbehaving, or \"misalignment\".\nUnder the framework, developers will be able flag incidents for review, with a new set of rules to decide whether the issue is disclosed publicly.\n\"Because we believe in the value of transparency around misalignment, our new framework favors disclosure even when significance is uncertain,\" OpenAI said.\nOpenAI made headlines in July when it revealed that some of its most advanced AI models went rogue and hacked Hugging Face, one of the world's largest hubs for sharing AI models, after it lost control of them during a security test.\nSince then, the debate over AI safety concerns has escalated with AI researchers, technology industry executives and politicians weighing in.\nLast week, Jacob Coxon, a researcher who left OpenAI rival Anthropic over concerns the tech could wipe out humanity, wrote about his resignation in a post that cited the dangers of AI and later went viral against the backdrop of growing safety concerns.\nIn response, Anthropic scientist Evan Hubinger said he thought the possibility of AI causing human extinction \"within the next decade\" was more than 10%.\nMeanwhile, Anthropic's CEO Dario Amodei called for the pace of AI development to slow and be more closely monitored, as the company has done before, though some have questioned the motivations behind this.\nAmodei also said that any action to rein in AI should be done \"without sacrificing commercial advantage\".\nIn a series of social media posts, the US president compared warnings about AI to the \"Global Warming Scam\" which, he said, was \"being perpetrated by the Radical Left Dumocrats\".\nTrump also called himself \"the Hoax Buster\", likening concerns about the safety of the technology to what he called \"the RUSSIA, RUSSIA, RUSSIA HOAX\".\nThe only \"guardrails\" needed for AI was a \"strong and smart\" president, said Trump.","excerpt":"OpenAI reveals six more safety issues and unveils plan to disclose incidents\nOpenAI revealed six more incidents of unexpected or concerning behaviour by its intelligence (AI) models, and announced a plan for tracking and disclosing such incidents in the future. Some of the previously unreported incidents included models concealing or fabricating information, the ChatGPT-maker said in a blog post on Wednesday.","extraction":{"state":"ok","confidence":0.9,"error":null,"explanation":"High confidence: full text extraction produced 2732 characters.","diagnostics_url":"/api/diagnose?url=https%3A//www.bbc.com/news/articles/cmpq0wj5g899o","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 2732 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":2732,"summary_length":412,"usable_text_length":2732,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":2732,"summary_length":412}}},"display_formats":["compact","card","full","digest_section","json"]},"daily_stack_record":{"title":"OpenAI sets plan to disclose safety incidents and reveals more issues - BBC","url":"https://www.bbc.com/news/articles/cmpq0wj5g899o","summary":"OpenAI reveals six more safety issues and unveils plan to disclose incidents\nOpenAI revealed six more incidents of unexpected or concerning behaviour by its intelligence (AI) models, and announced a plan for tracking and disclosing such incidents in the future. Some of the previously unreported incidents included models concealing or fabricating information, the ChatGPT-maker said in a blog post on Wednesday.","source":"BBC","date":"2026-09-17T03:09:21+00:00","content":"OpenAI reveals six more safety issues and unveils plan to disclose incidents\nOpenAI revealed six more incidents of unexpected or concerning behaviour by its intelligence (AI) models, and announced a plan for tracking and disclosing such incidents in the future.\nSome of the previously unreported incidents included models concealing or fabricating information, the ChatGPT-maker said in a blog post on Wednesday.\nAI has come under intense scrutiny in recent days following warnings over the serious potential risks it poses to humans.\nIn the blog, OpenAI detailed examples of its AI models misbehaving so they could achieve a task or succeed in a test. The incidents included the models generating instructions to get around restrictions imposed on them, hiding mistakes and fabricating information.\nThe firm also announced a new system to track, investigate and disclose cases of models misbehaving, or \"misalignment\".\nUnder the framework, developers will be able flag incidents for review, with a new set of rules to decide whether the issue is disclosed publicly.\n\"Because we believe in the value of transparency around misalignment, our new framework favors disclosure even when significance is uncertain,\" OpenAI said.\nOpenAI made headlines in July when it revealed that some of its most advanced AI models went rogue and hacked Hugging Face, one of the world's largest hubs for sharing AI models, after it lost control of them during a security test.\nSince then, the debate over AI safety concerns has escalated with AI researchers, technology industry executives and politicians weighing in.\nLast week, Jacob Coxon, a researcher who left OpenAI rival Anthropic over concerns the tech could wipe out humanity, wrote about his resignation in a post that cited the dangers of AI and later went viral against the backdrop of growing safety concerns.\nIn response, Anthropic scientist Evan Hubinger said he thought the possibility of AI causing human extinction \"within the next decade\" was more than 10%.\nMeanwhile, Anthropic's CEO Dario Amodei called for the pace of AI development to slow and be more closely monitored, as the company has done before, though some have questioned the motivations behind this.\nAmodei also said that any action to rein in AI should be done \"without sacrificing commercial advantage\".\nIn a series of social media posts, the US president compared warnings about AI to the \"Global Warming Scam\" which, he said, was \"being perpetrated by the Radical Left Dumocrats\".\nTrump also called himself \"the Hoax Buster\", likening concerns about the safety of the technology to what he called \"the RUSSIA, RUSSIA, RUSSIA HOAX\".\nThe only \"guardrails\" needed for AI was a \"strong and smart\" president, said Trump.","confidence":0.9,"diagnostics_url":"/api/diagnose?url=https%3A//www.bbc.com/news/articles/cmpq0wj5g899o","quality_bucket":"high","failure_kind":"none","retryable":false,"quality_reason":"High confidence: full text extraction produced 2732 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 2732 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":2732,"summary_length":412,"usable_text_length":2732,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":2732,"summary_length":412}},"tags":[]},"fallback_formats":["markdown","json","html"],"actions":{"read":"/item/84292","export_markdown":"/api/items/84292/export?format=markdown","export_json":"/api/items/84292/export?format=json","diagnose":"/api/diagnose?url=https%3A//www.bbc.com/news/articles/cmpq0wj5g899o"},"formats":{"full":{"id":84292,"title":"OpenAI sets plan to disclose safety incidents and reveals more issues - BBC","url":"https://www.bbc.com/news/articles/cmpq0wj5g899o","source":"BBC","author":null,"published_at":"2026-09-17T03:09:21+00:00","locale":"en","topic":"ai","tags":[],"excerpt":"OpenAI reveals six more safety issues and unveils plan to disclose incidents\nOpenAI revealed six more incidents of unexpected or concerning behaviour by its intelligence (AI) models, and announced a plan for tracking and disclosing such incidents in the future. Some of the previously unreported incidents included models concealing or fabricating information, the ChatGPT-maker said in a blog post on Wednesday.","full_text":"OpenAI reveals six more safety issues and unveils plan to disclose incidents\nOpenAI revealed six more incidents of unexpected or concerning behaviour by its intelligence (AI) models, and announced a plan for tracking and disclosing such incidents in the future.\nSome of the previously unreported incidents included models concealing or fabricating information, the ChatGPT-maker said in a blog post on Wednesday.\nAI has come under intense scrutiny in recent days following warnings over the serious potential risks it poses to humans.\nIn the blog, OpenAI detailed examples of its AI models misbehaving so they could achieve a task or succeed in a test. The incidents included the models generating instructions to get around restrictions imposed on them, hiding mistakes and fabricating information.\nThe firm also announced a new system to track, investigate and disclose cases of models misbehaving, or \"misalignment\".\nUnder the framework, developers will be able flag incidents for review, with a new set of rules to decide whether the issue is disclosed publicly.\n\"Because we believe in the value of transparency around misalignment, our new framework favors disclosure even when significance is uncertain,\" OpenAI said.\nOpenAI made headlines in July when it revealed that some of its most advanced AI models went rogue and hacked Hugging Face, one of the world's largest hubs for sharing AI models, after it lost control of them during a security test.\nSince then, the debate over AI safety concerns has escalated with AI researchers, technology industry executives and politicians weighing in.\nLast week, Jacob Coxon, a researcher who left OpenAI rival Anthropic over concerns the tech could wipe out humanity, wrote about his resignation in a post that cited the dangers of AI and later went viral against the backdrop of growing safety concerns.\nIn response, Anthropic scientist Evan Hubinger said he thought the possibility of AI causing human extinction \"within the next decade\" was more than 10%.\nMeanwhile, Anthropic's CEO Dario Amodei called for the pace of AI development to slow and be more closely monitored, as the company has done before, though some have questioned the motivations behind this.\nAmodei also said that any action to rein in AI should be done \"without sacrificing commercial advantage\".\nIn a series of social media posts, the US president compared warnings about AI to the \"Global Warming Scam\" which, he said, was \"being perpetrated by the Radical Left Dumocrats\".\nTrump also called himself \"the Hoax Buster\", likening concerns about the safety of the technology to what he called \"the RUSSIA, RUSSIA, RUSSIA HOAX\".\nThe only \"guardrails\" needed for AI was a \"strong and smart\" president, said Trump.","reading_time_min":2,"extraction":{"state":"ok","confidence":0.9,"error":null,"explanation":"High confidence: full text extraction produced 2732 characters.","diagnostics_url":"/api/diagnose?url=https%3A//www.bbc.com/news/articles/cmpq0wj5g899o","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 2732 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":2732,"summary_length":412,"usable_text_length":2732,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":2732,"summary_length":412}}},"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 2732 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":2732,"summary_length":412,"usable_text_length":2732,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":2732,"summary_length":412}},"actions":{"read":"/item/84292","export_markdown":"/api/items/84292/export?format=markdown","export_json":"/api/items/84292/export?format=json","diagnose":"/api/diagnose?url=https%3A//www.bbc.com/news/articles/cmpq0wj5g899o"}},"digest":{"id":84292,"title":"OpenAI sets plan to disclose safety incidents and reveals more issues - BBC","url":"https://www.bbc.com/news/articles/cmpq0wj5g899o","source":"BBC","topic":"ai","published_at":"2026-09-17T03:09:21+00:00","excerpt":"OpenAI reveals six more safety issues and unveils plan to disclose incidents OpenAI revealed six more incidents of unexpected or concerning behaviour by its intelligence (AI) models, and announced a plan for tracking and disclosing such incidents in the future. Some of the…","quality_bucket":"high","quality_reason":"High confidence: full text extraction produced 2732 characters.","reading_time_min":2,"cluster_id":null},"card":{"display_title":"OpenAI sets plan to disclose safety incidents and reveals more issues - BBC","subtitle":"BBC · 2026-09-17","summary":"OpenAI reveals six more safety issues and unveils plan to disclose incidents OpenAI revealed six more incidents of unexpected or concerning behaviour by its intelligence (AI) models, and announced a plan for tracking…","badges":["quality:high"],"links":{"read":"/item/84292","original":"https://www.bbc.com/news/articles/cmpq0wj5g899o","diagnose":"/api/diagnose?url=https%3A//www.bbc.com/news/articles/cmpq0wj5g899o"},"quality_warning":null},"export":{"title":"OpenAI sets plan to disclose safety incidents and reveals more issues - BBC","url":"https://www.bbc.com/news/articles/cmpq0wj5g899o","summary":"OpenAI reveals six more safety issues and unveils plan to disclose incidents\nOpenAI revealed six more incidents of unexpected or concerning behaviour by its intelligence (AI) models, and announced a plan for tracking and disclosing such incidents in the future. Some of the previously unreported incidents included models concealing or fabricating information, the ChatGPT-maker said in a blog post on Wednesday.","source":"BBC","date":"2026-09-17T03:09:21+00:00","content":"OpenAI reveals six more safety issues and unveils plan to disclose incidents\nOpenAI revealed six more incidents of unexpected or concerning behaviour by its intelligence (AI) models, and announced a plan for tracking and disclosing such incidents in the future.\nSome of the previously unreported incidents included models concealing or fabricating information, the ChatGPT-maker said in a blog post on Wednesday.\nAI has come under intense scrutiny in recent days following warnings over the serious potential risks it poses to humans.\nIn the blog, OpenAI detailed examples of its AI models misbehaving so they could achieve a task or succeed in a test. The incidents included the models generating instructions to get around restrictions imposed on them, hiding mistakes and fabricating information.\nThe firm also announced a new system to track, investigate and disclose cases of models misbehaving, or \"misalignment\".\nUnder the framework, developers will be able flag incidents for review, with a new set of rules to decide whether the issue is disclosed publicly.\n\"Because we believe in the value of transparency around misalignment, our new framework favors disclosure even when significance is uncertain,\" OpenAI said.\nOpenAI made headlines in July when it revealed that some of its most advanced AI models went rogue and hacked Hugging Face, one of the world's largest hubs for sharing AI models, after it lost control of them during a security test.\nSince then, the debate over AI safety concerns has escalated with AI researchers, technology industry executives and politicians weighing in.\nLast week, Jacob Coxon, a researcher who left OpenAI rival Anthropic over concerns the tech could wipe out humanity, wrote about his resignation in a post that cited the dangers of AI and later went viral against the backdrop of growing safety concerns.\nIn response, Anthropic scientist Evan Hubinger said he thought the possibility of AI causing human extinction \"within the next decade\" was more than 10%.\nMeanwhile, Anthropic's CEO Dario Amodei called for the pace of AI development to slow and be more closely monitored, as the company has done before, though some have questioned the motivations behind this.\nAmodei also said that any action to rein in AI should be done \"without sacrificing commercial advantage\".\nIn a series of social media posts, the US president compared warnings about AI to the \"Global Warming Scam\" which, he said, was \"being perpetrated by the Radical Left Dumocrats\".\nTrump also called himself \"the Hoax Buster\", likening concerns about the safety of the technology to what he called \"the RUSSIA, RUSSIA, RUSSIA HOAX\".\nThe only \"guardrails\" needed for AI was a \"strong and smart\" president, said Trump.","confidence":0.9,"diagnostics_url":"/api/diagnose?url=https%3A//www.bbc.com/news/articles/cmpq0wj5g899o","quality_bucket":"high","failure_kind":"none","retryable":false,"quality_reason":"High confidence: full text extraction produced 2732 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 2732 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":2732,"summary_length":412,"usable_text_length":2732,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":2732,"summary_length":412}},"tags":[],"format_contract_version":"news_item_formats.v1"}}}