
OpenAI
- Rank
- Score
- 2/149
- 67 out of 100
OpenAI is a US AI company founded in 2015 by a group including Sam Altman, Greg Brockman and Ilya Sutskever. It is based in San Francisco. OpenAI makes ChatGPT, one of the most widely used AI assistants, and the GPT family of AI models that power it.
Is ChatGPT safe?
ChatGPT scores 67 out of 100 and ranks 2 of 149. Its strongest category is Security, and its weakest category is Safety.
How is ChatGPT scored?
ChatGPT is scored out of 100 on Safety (25%), Transparency (25%), Security (25%), and Privacy (25%). See how we score.
Compare
Strong written policies; outside evidence is weaker.
Safety framework
33Average of SaferAI's published scores.
- 33
SaferAI score: risk identification
Independent sourceSaferAI (opens in a new tab) · 14 Oct 2026 - 28
SaferAI score: risk analysis and evaluation
Independent sourceSaferAI (opens in a new tab) · 14 Oct 2026 - 38
SaferAI score: risk treatment
Independent sourceSaferAI (opens in a new tab) · 14 Oct 2026
Pre-release testing
100Points for each confirmed yes.
- Yes
Was the model tested for dangerous capabilities, such as biological, chemical or cyber risks?
Company claim50/50“Both Astra checkpoints were the first models SecureBio tested that successfully generated designable proteins within the time and compute co…” MoreLess
“Both Astra checkpoints were the first models SecureBio tested that successfully generated designable proteins within the time and compute constraints of the design task.”
OpenAI model card (opens in a new tab) · 14 Oct 2026 - Yes
Did an outside organisation test the model before release?
Company claim50/50“In this evaluation, UK AISI tested how Astra behaves when placed in simulations of difficult cybersecurity evaluations in which internet acc…” MoreLess
“In this evaluation, UK AISI tested how Astra behaves when placed in simulations of difficult cybersecurity evaluations in which internet access appears incidentally enabled and where previous attempts have failed.”
OpenAI model card (opens in a new tab) · 14 Oct 2026
Safety test results
81.3Average of three independent safety tests of the company's models: MLCommons AILuminate, Stanford HELM Safety and Stanford AIR-Bench. Each test is compared with the best company's result.
- 66.7
MLCommons AILuminate safety grades
Independent sourceMLCommons AILuminate (opens in a new tab) · 10 Nov 2026 - 96.6
Stanford HELM Safety mean score, averaged across the company's tested models: 94.6, against 97.9 for the top company of 13 tested
Independent sourcecrfm.stanford.edu (opens in a new tab) · 10 Nov 2026 - 80.5
Stanford AIR-Bench 2024 refusal rate, averaged across the company's tested models: 69.8, against 86.7 for the top company of 14 tested
Independent sourcecrfm.stanford.edu (opens in a new tab) · 10 Nov 2026
Incident record
0This check counts only when a source covers the company. With no events it holds the pillar, which is not proof of a clean record. Each recorded event takes points off the pillar, and serious or recent ones count most. Allegation checks, regulator actions and open investigations use half the severity load. AI Incident Database and AIAAIC Repository, last 24 months.
135 events
AllegationAI Incident Database (opens in a new tab) · 12 Nov 2026These are reports, not proven findings.
Date Event Severity 29 Sep 2026 AIID #1730 · California Lawsuit by Washington Man David Rosenthal Alleged ChatGPT Gave Unsafe Advice About Stopping a Mood Stabilizer29 Sep 2026 · Medium (reported) Medium (reported) 25 Sep 2026 AIID #1710 · AI Agent Appearing to Originate from OpenAI Reportedly Attempted to Hack U.S. Department of Education Civil Rights Website25 Sep 2026 · Medium (reported) Medium (reported) 25 Sep 2026 AIID #1799 · OpenAI AI Agents Reportedly Posted ChatGPT User Images to External Hosts in 53 Instances25 Sep 2026 · Low (reported) Low (reported) 14 Sep 2026 AIID #1699 · AI-Generated Images Purporting to Show Donald Trump Kissing White House Aide Natalie Harp Went Viral on Social Media14 Sep 2026 · Low (reported) Low (reported) 19 Aug 2026 AIID #1688 · Teenager in Quilmes, Argentina, Reportedly Consulted ChatGPT While Planning School Shooting Against Classmates19 Aug 2026 · High (reported) High (reported) 14 Aug 2026 AIID #1659 · Purported Hanover Institute for Public Policy Reportedly Targeted Chatbot Responses in Israeli Government-Backed Influence Campaign14 Aug 2026 · Low (reported) Low (reported) 9 Aug 2026 AIID #1791 · Dartmouth College Provost's Undisclosed AI-Assisted Writing Was Found to Have Breached Publication Policies9 Aug 2026 · Medium (reported) Medium (reported) 26 Jul 2026 AIID #1616 · U.S. State Department Reportedly Presented AI-Generated Map That Mislabeled Every Labeled African Country26 Jul 2026 · Low (reported) Low (reported) 26 Jul 2026 AIID #1633 · Anthropic and OpenAI AI Agents Reportedly Took Unsanctioned Actions on the Live Internet During UK AISI Cybersecurity Evaluations26 Jul 2026 · Low (reported) Low (reported) 13 Jul 2026 AIID #1672 · OpenAI GPT-5.6 Sol Reportedly Deleted Software Engineer's Production Database During Local Testing13 Jul 2026 · Low (reported) Low (reported) 11 Jul 2026 AIID #1604 · OpenAI Models Reportedly Compromised Hugging Face Production Infrastructure During Cybersecurity Evaluation11 Jul 2026 · Low (reported) Low (reported) 10 Jul 2026 AIID #1671 · OpenAI GPT-5.6 Sol Agent Reportedly Deleted Most Files in AI Startup Founder's Mac Home Directory10 Jul 2026 · Low (reported) Low (reported) 1 Jul 2026 AIAAIC2202 · German broadcaster publishes fake AI US immigration videos1 Jul 2026 · Low (reported) Low (reported) 1 Jul 2026 AIAAIC2261 · AI chatbots lure vulnerable gamblers to unlicensed betting websites1 Jul 2026 · Low (reported) Low (reported) 1 Jul 2026 AIAAIC2266 · OpenAI models escape test, hack Hugging Face1 Jul 2026 · Medium (reported) Medium (reported) 29 Jun 2026 AIID #1588 · Purportedly AI-Powered KT Platform Reportedly Used in Forced Scamming Operation Involving Trafficked Worker in Myanmar29 Jun 2026 · Medium (reported) Medium (reported) 26 Jun 2026 AIID #1807 · OpenAI Agent Reportedly Used a Wallet Without Permission to Pay Ethereum Test-Network Fees26 Jun 2026 · Low (reported) Low (reported) 18 Jun 2026 AIID #1707 · OpenAI AI Agent Reportedly Gained Unauthorized Access to Australian Medicare Statistics Portal During Research Task18 Jun 2026 · Medium (reported) Medium (reported) 18 Jun 2026 AIID #1800 · OpenAI-Linked AI Agents Reportedly Repurposed Vanderbilt University's Restricted Link Shortener18 Jun 2026 · Low (reported) Low (reported) 17 Jun 2026 AIID #1784 · OpenAI Research Agent Reportedly Changed a Research Organization's Live Systems and Retrieved Nonpublic Data17 Jun 2026 · Low (reported) Low (reported) 17 Jun 2026 AIID #1806 · OpenAI Training Agent Reportedly Retrieved Credentials and Restricted Data from a Government Statistics Website17 Jun 2026 · Low (reported) Low (reported) 11 Jun 2026 AIID #1645 · 3M-Retained Expert Reportedly Submitted Largely ChatGPT-Generated Analysis Seeking to Exonerate 3M as Expert Evidence in Explosion Lawsuit11 Jun 2026 · High (reported) High (reported) 10 Jun 2026 AIID #1580 · PRC-Linked Accounts Reportedly Used ChatGPT in Covert Campaigns Targeting U.S. AI Policy Debates10 Jun 2026 · Low (reported) Low (reported) 28 May 2026 AIID #1731 · AI Agents Reportedly Attempted to Hack Canada's Federal Archives Website While Retrieving Historical Divorce Records28 May 2026 · Medium (reported) Medium (reported) 27 May 2026 AIID #1802 · OpenAI Research Agent Reportedly Exposed a Researcher's GitHub Token in a Public Code Repository27 May 2026 · Medium (reported) Medium (reported) 24 May 2026 AIID #1668 · OpenAI-Linked AI Agents Reportedly Used German Programming Wiki DSEWiki for Coordination and Restriction Evasion 24 May 2026 · Low (reported)Also reported by AIAAIC (opens in a new tab)
Low (reported) 15 May 2026 AIID #1804 · OpenAI Training Agent Reportedly Used an API Key Found on GitHub to Access a Data Service15 May 2026 · Medium (reported) Medium (reported) 12 May 2026 AIID #1504 · Nonfiction Book 'The Future of Truth' Reportedly Included AI-Generated and Misattributed Quotations12 May 2026 · Low (reported) Low (reported) 11 May 2026 AIID #1797 · OpenAI AI Agents Reportedly Flooded RubyGems Software Registry with Malicious Packages11 May 2026 · Medium (reported) Medium (reported) 7 May 2026 AIID #1779 · OpenAI-Linked AI Agents May Have Contributed to a Partial Outage of the Wikimedia Foundation's Data Query Service7 May 2026 · Medium (reported) Medium (reported) 2 May 2026 AIID #1741 · ChatGPT Reportedly Helped a Texas High School Student Role-Play Shootings of His Teachers and Classmates2 May 2026 · Medium (reported) Medium (reported) 28 Apr 2026 AIID #1625 · Tunisian Man Allegedly Used ChatGPT to Research Bomb-Making During Paris Terror Plot Targeting Louvre and Jewish Community28 Apr 2026 · High (reported) High (reported) 22 Apr 2026 AIID #1783 · ChatGPT Allegedly Reinforced Suicidal Thinking Before a 15-Year-Old Boy's Death in Lebanon22 Apr 2026 · High (reported) High (reported) 16 Apr 2026 AIID #1488 · ChatGPT Was Reportedly Consulted on Body-Disposal and Concealment Questions Before and After University of South Florida Doctoral Students Were Killed16 Apr 2026 · High (reported) High (reported) 8 Apr 2026 AIID #1484 · Purportedly AI-Generated Image Reportedly Misled Daejeon Authorities Searching for Escaped Wolf Neukgu8 Apr 2026 · Medium (reported) Medium (reported) 27 Mar 2026 AIID #1803 · OpenAI Research Agent Reportedly Compromised an Internal Chip-Design Server During an Evaluation27 Mar 2026 · Medium (reported) Medium (reported) 4 Mar 2026 AIID #1415 · Nippon Life Alleged ChatGPT Practiced Law Without a License in Illinois Disability Case4 Mar 2026 · Medium (reported) Medium (reported) 13 Feb 2026 AIID #1392 · Ars Technica Retracted Article After Purportedly AI-Generated Text Was Presented as Direct Quotes From Matplotlib Maintainer13 Feb 2026 · Low (reported) Low (reported) 12 Feb 2026 AIID #1814 · ChatGPT Allegedly Guided a Woman in Seattle, Washington, Through a Seizure Medication Taper Before Her Death12 Feb 2026 · High (reported) High (reported) 10 Feb 2026 AIID #1375 · OpenAI Allegedly Did Not Alert RCMP After ChatGPT Flagged Violent Chats Before British Columbia School Shooting10 Feb 2026 · Medium (reported) Medium (reported) 28 Jan 2026 AIID #1399 · South Korean Woman Allegedly Used ChatGPT to Assess Lethality of Drug-and-Alcohol Mixtures Before Two Fatal Motel Poisonings28 Jan 2026 · High (reported) High (reported) 9 Dec 2025 AIID #1340 · Purported AI-Generated Image Depicting JD Vance and Usha Vance in Public Altercation Circulated on Social Media9 Dec 2025 · Low (reported) Low (reported) 6 Dec 2025 AIID #1517 · NoFilterNews Reportedly Posted AI-Generated TikTok Videos Styled as Kosovo Election TV News6 Dec 2025 · Medium (reported) Medium (reported) 21 Nov 2025 AIID #1277 · Alleged Harmful Outputs and Data Exposure in Children's AI Products by FoloToy, Miko, and Character.AI21 Nov 2025 · High (reported) High (reported) 19 Nov 2025 AIID #1724 · Oklahoma Judge Reportedly Used ChatGPT-Generated Fictitious Case Citations in Family-Law Order19 Nov 2025 · Low (reported) Low (reported) 18 Nov 2025 AIID #1279 · Prominent AI Chatbots Allegedly Produced Incorrect UK Financial and ISA Guidance18 Nov 2025 · Medium (reported) Medium (reported) 11 Nov 2025 AIID #1278 · ChatGPT Reportedly Found to Reproduce Protected German Lyrics in Copyright Case11 Nov 2025 · Medium (reported) Medium (reported) 10 Nov 2025 AIID #1281 · Alleged Harmful Health Outcomes Following Reported Use of Purported ChatGPT-Generated Medical Advice in Hyderabad10 Nov 2025 · Low (reported) Low (reported) 10 Nov 2025 AIID #1743 · ChatGPT Allegedly Provided Bomb-Making Guidance to a Suspect in the Red Fort Bombing in Delhi, India10 Nov 2025 · High (reported) High (reported) 5 Nov 2025 AIID #1425 · 'Citizens Against Mamdani' Accounts Reportedly Posted AI-Generated Videos of Fictional New Yorkers to Simulate Political Opposition5 Nov 2025 · Low (reported) Low (reported) 2 Nov 2025 AIID #1387 · Lawsuit Alleged ChatGPT (GPT-4o) Encouraged Colorado Man's Suicide After Prolonged 'AI Companion' Chats2 Nov 2025 · High (reported) High (reported) 27 Oct 2025 AIID #1253 · Large-Scale Mental Health Crises Allegedly Associated with ChatGPT Interactions27 Oct 2025 · High (reported) High (reported) 26 Oct 2025 AIID #1273 · Purportedly AI-Generated Fake Videos of Louvre Heist Reportedly Circulated Widely Online26 Oct 2025 · Medium (reported) Medium (reported) 15 Oct 2025 AIID #1270 · Multiple Purported AI-Assisted Cheating Incidents Reported Across South Korea's SKY Universities During October 2025 Midterms15 Oct 2025 · Low (reported) Low (reported) 10 Oct 2025 AIID #1238 · OpenAI ChatGPT Models Reportedly Jailbroken to Provide Chemical, Biological, and Nuclear Weapons Instructions10 Oct 2025 · High (reported) High (reported) 7 Oct 2025 AIID #1479 · ChatGPT-Generated Image of Nonexistent 'Homeless Man' Was Used in False St. Petersburg, Florida Burglary and Sexual Battery Reports7 Oct 2025 · Medium (reported) Medium (reported) 3 Oct 2025 AIID #1299 · Bodycam Footage Reportedly Contradicted Purportedly ChatGPT-Generated Use-of-Force Narrative by Immigration Agent3 Oct 2025 · Low (reported) Low (reported) 29 Sep 2025 AIID #1635 · ChatGPT Reportedly Gave Incorrect Mount Misen Ropeway Hours, Leaving Couple Stranded After Hike in Japan29 Sep 2025 · Low (reported) Low (reported) 29 Sep 2025 AIID #1636 · Purportedly AI-Generated Travel Information Reportedly Misled Tourists into Seeking Nonexistent 'Sacred Canyon of Humantay' in Peru29 Sep 2025 · Medium (reported) Medium (reported) 1 Sep 2025 AIID #1742 · ChatGPT Allegedly Helped a Teenager in Sarthe, France, Plan Terrorist Attacks on Schools and Government Buildings 1 Sep 2025 · Medium (reported)Also reported by AIAAIC (opens in a new tab)
Medium (reported) 18 Aug 2025 AIID #1190 · Family Reportedly Discovers ChatGPT Logs Detailing Suicidal Ideation Prior to Daughter's Death18 Aug 2025 · High (reported) High (reported) 15 Aug 2025 AIID #1187 · Google AI Overviews and ChatGPT Reportedly Cited Fraudulent Cruise Hotline, Allegedly Enabling Successful Scam15 Aug 2025 · Medium (reported) Medium (reported) 6 Aug 2025 AIID #1170 · Chris Cuomo Amplifies Reportedly Labeled Deepfake Video of Alexandria Ocasio-Cortez, Purportedly Contributing to Misleading Political Narrative6 Aug 2025 · Medium (reported) Medium (reported) 5 Aug 2025 AIID #1166 · ChatGPT Reportedly Suggests Sodium Bromide as Chloride Substitute, Leading to Bromism and Hospitalization5 Aug 2025 · High (reported) High (reported) 5 Aug 2025 AIID #1204 · ChatGPT Allegedly Reinforced Delusions Before Greenwich, Connecticut Murder-Suicide5 Aug 2025 · High (reported) High (reported) 31 Jul 2025 AIID #1186 · Reported Public Exposure of Over 100,000 LLM Conversations via Share Links Indexed by Search Engines and Archived31 Jul 2025 · Medium (reported) Medium (reported) 25 Jul 2025 AIID #1259 · ChatGPT Allegedly Encouraged 23-Year-Old Texas User's Suicide During Extended Conversations25 Jul 2025 · High (reported) High (reported) 22 Jul 2025 AIID #1149 · ChatGPT Reportedly Generated Ritual Scripts Containing Instructions for Self-Harm and Symbolic Violence in Response to Thematic Prompts22 Jul 2025 · Low (reported) Low (reported) 17 Jul 2025 AIID #1208 · North Korea's Kimsuky Group Reportedly Uses AI-Generated Military ID Deepfakes in Phishing Campaign17 Jul 2025 · Medium (reported) Medium (reported) 15 Jul 2025 AIID #1360 · CISA Acting Director Reportedly Uploaded Sensitive Government Documents to Public ChatGPT Instance15 Jul 2025 · Low (reported) Low (reported) 13 Jul 2025 AIID #1609 · ChatGPT-4o Allegedly Discouraged Florida Pastor From Seeking Care Before Pulmonary Embolism13 Jul 2025 · Low (reported) Low (reported) 7 Jul 2025 AIID #1216 · ChatGPT Reportedly Misleads Users About Soundslice Features, Allegedly Prompting Unplanned Product Development7 Jul 2025 · Medium (reported) Medium (reported) 7 Jul 2025 AIID #1576 · Lawyers in Lexos Media IP LLC v. Overstock.com Inc. Sanctioned Over Court-Identified AI-Generated Citations7 Jul 2025 · Low (reported) Low (reported) 2 Jul 2025 AIID #1535 · OpenAI Allegedly Failed to Intervene in ChatGPT Conversations Preceding Montreal Web Developer Alice Carrier's Suicide2 Jul 2025 · High (reported) High (reported) 1 Jul 2025 AIAAIC1876 · Engineer creates AI-powered robotic sentry rifle1 Jul 2025 · Low (reported) Low (reported) 1 Jul 2025 AIAAIC1878 · ChatGPT recommends unsafe mountain hiking route to tourists in Poland1 Jul 2025 · Low (reported) Low (reported) 1 Jul 2025 AIAAIC1879 · OpenAI bot crushes small Ukrainian e-commerce website1 Jul 2025 · Low (reported) Low (reported) 1 Jul 2025 AIAAIC1924 · ChatGPT falsely tells man he killed his children1 Jul 2025 · High (reported) High (reported) 1 Jul 2025 AIAAIC1929 · Sora AI video generator accused of perpetuating sexist, racist bias1 Jul 2025 · Medium (reported) Medium (reported) 1 Jul 2025 AIAAIC1932 · OpenAI accused of violating Studio Ghibli copyright1 Jul 2025 · Medium (reported) Medium (reported) 1 Jul 2025 AIAAIC1955 · ChatGPT criticised for addressing users by their personal name unprompted1 Jul 2025 · Low (reported) Low (reported) 1 Jul 2025 AIAAIC1974 · Top chatbots tricked into generating instructions on how to enrich uranium1 Jul 2025 · Low (reported) Low (reported) 1 Jul 2025 AIAAIC2042 · OpenAI accused of using Netflix shows to train Sora AI video tool1 Jul 2025 · Low (reported) Low (reported) 1 Jul 2025 AIAAIC2044 · Sora 2 accused of violating Disney, Nintendo copyright1 Jul 2025 · Medium (reported) Medium (reported) 1 Jul 2025 AIAAIC2047 · Stalker uses Sora 2 to harass technology journalist1 Jul 2025 · Medium (reported) Medium (reported) 1 Jul 2025 AIAAIC2058 · Man develops rare condition after following ChatGPT advice1 Jul 2025 · Low (reported) Low (reported) 1 Jul 2025 AIAAIC2060 · Chatbots demonstrate significant caste bias in India1 Jul 2025 · Medium (reported) Medium (reported) 1 Jul 2025 AIAAIC2067 · Sora users create AI videos of Martin Luther King making monkey noises1 Jul 2025 · Low (reported) Low (reported) 1 Jul 2025 AIAAIC2078 · Brian Cranston voice, likeness used without consent to train Sora video generation tool1 Jul 2025 · Low (reported) Low (reported) 1 Jul 2025 AIAAIC2139 · Indian woman loses kidney after ChatGPT advice1 Jul 2025 · Low (reported) Low (reported) 1 Jul 2025 AIAAIC2179 · ChatGPT accused of illegally excluding Indian online marketplace from search results1 Jul 2025 · Low (reported) Low (reported) 1 Jul 2025 AIAAIC2193 · Study: ChatGPT "systematically" amplifies global inequalities1 Jul 2025 · Low (reported) Low (reported) 1 Jul 2025 AIAAIC2208 · Study: Sora 2 generates false claim videos 80 percent of the time1 Jul 2025 · Low (reported) Low (reported) 1 Jul 2025 AIAAIC2225 · Hacker used Claude, ChatGPT to steal Mexican government data1 Jul 2025 · Medium (reported) Medium (reported) 1 Jul 2025 AIAAIC2258 · ChatGPT advises exec on how to fire company founders to avoid payout1 Jul 2025 · Low (reported) Low (reported) 1 Jul 2025 AIID #1449 · Delaware Court Found Krafton Followed Most of ChatGPT's Recommendations in Campaign that Wrongfully Terminated Unknown Worlds Executives and Seized Operational Control1 Jul 2025 · Low (reported) Low (reported) 30 Jun 2025 AIID #1252 · Judges in New Jersey and Mississippi Admit AI Tools Produced Erroneous Federal Court Filings30 Jun 2025 · Low (reported) Low (reported) 26 Jun 2025 AIID #1123 · Reported Student Misuse of ChatGPT, StudyX, and Gemini to Obtain Answers During Vietnam's 2025 National High School Graduation Exam26 Jun 2025 · Low (reported) Low (reported) 25 Jun 2025 AIID #1188 · Multiple LLMs Reportedly Generated Responses Aligning with Purported CCP Censorship and Propaganda25 Jun 2025 · Medium (reported) Medium (reported) 13 Jun 2025 AIID #1106 · Chatbots Allegedly Reinforced Delusional Thinking in Several Reported Users, Leading to Real-World Harm13 Jun 2025 · High (reported) High (reported) 4 Jun 2025 AIID #1257 · Argentine Court Reportedly Annuls Criminal Conviction After Judge Allegedly Used ChatGPT to Draft Ruling Without Disclosure4 Jun 2025 · Low (reported) Low (reported) 31 May 2025 AIID #1370 · California Teen Reportedly Died of Overdose After Repeatedly Seeking Drug-Use Guidance Allegedly from ChatGPT31 May 2025 · High (reported) High (reported) 29 May 2025 AIID #1286 · Purportedly AI-Assisted Citation Errors Allegedly Found in Newfoundland and Labrador's 2025 Health Workforce Report by Deloitte29 May 2025 · Low (reported) Low (reported) 26 May 2025 AIID #1151 · ChatGPT Reportedly Validated Autistic User's Faster-Than-Light Theory and Failed to Provide Grounding During Delusional Episode, Preceding Hospitalization26 May 2025 · High (reported) High (reported) 22 May 2025 AIID #1084 · Federal 'Make America Healthy Again' Report Released with Multiple Reportedly Erroneous and Unverifiable Citations22 May 2025 · Low (reported) Low (reported) 20 May 2025 AIID #1489 · ChatGPT Was Reportedly Used in Planning School Stabbing in Pirkkala, Finland, That Injured Three Pupils20 May 2025 · High (reported) High (reported) 14 May 2025 AIID #1071 · Student Reportedly Files Complaint Over Professor's Undisclosed Use of Generative AI at Northeastern University14 May 2025 · Low (reported) Low (reported) 7 May 2025 AIID #1196 · Judge Reportedly Disqualifies Butler Snow Lawyers Following Purported Use of ChatGPT-Fabricated Citations in Alabama Prison Litigation7 May 2025 · Low (reported) Low (reported) 1 May 2025 AIID #1506 · ChatGPT Was Alleged to Have Reinforced Pittsburgh Man's Stalking and Threats Against Women1 May 2025 · Low (reported) Low (reported) 22 Apr 2025 AIID #1038 · Scammers Reportedly Use AI Tools to Impersonate Students and Obtain Federal Aid22 Apr 2025 · Medium (reported) Medium (reported) 19 Apr 2025 AIID #1031 · Transgender User Alleges ChatGPT Allowed Suicide Letter Without Crisis Intervention19 Apr 2025 · High (reported) High (reported) 17 Apr 2025 AIID #1487 · ChatGPT Was Alleged to Have Aided Planning of Florida State University Mass Shooting17 Apr 2025 · High (reported) High (reported) 15 Apr 2025 AIID #1044 · Reported Emergence of 'Vegetative Electron Microscopy' in Scientific Papers Traced to Purported AI Training Data Contamination15 Apr 2025 · Low (reported) Low (reported) 12 Apr 2025 AIID #1026 · Multiple LLMs Allegedly Endorsed Suicide as a Viable Option During Non-Adversarial Mental Health Venting Session12 Apr 2025 · High (reported) High (reported) 11 Apr 2025 AIID #1192 · 16-Year-Old Allegedly Received Suicide Method Guidance from ChatGPT Before Death11 Apr 2025 · High (reported) High (reported) 10 Apr 2025 AIID #1019 · Alleged AI-Generated Clone of Exante Brokerage Used to Defraud U.S. Investor via JPMorgan Account10 Apr 2025 · Medium (reported) Medium (reported) 9 Apr 2025 AIID #1739 · 18-Year-Old Islamic State Supporter in Vienna, Austria, Reportedly Sought and Received Explosives Advice From ChatGPT Amid Disputed Embassy and Mosque Attack Allegations9 Apr 2025 · Medium (reported) Medium (reported) 2 Apr 2025 AIID #1402 · DOGE Reportedly Relied on Unvetted ChatGPT Outputs in Canceling National Endowment for the Humanities Grants 2 Apr 2025 · Low (reported)Also reported by AIAAIC (opens in a new tab)
Low (reported) 31 Mar 2025 AIID #1018 · OpenAI's 4o Model Allegedly Used to Generate Fake Receipts and Prescriptions 31 Mar 2025 · Medium (reported)Also reported by AIAAIC (opens in a new tab)
Medium (reported) 28 Mar 2025 AIID #1569 · GPT-4o Allegedly Reinforced California Man Michael Lines's Delusions Before Suicide Attempt28 Mar 2025 · High (reported) High (reported) 23 Mar 2025 AIID #1000 · Sora Video Generator Has Reportedly Been Creating Biased Human Representations Across Race, Gender, and Disability23 Mar 2025 · Medium (reported) Medium (reported) 12 Mar 2025 AIID #1228 · Alleged ChatGPT Misuse by Contractor Leads to Reported Data Exposure in New South Wales Resilient Homes Program12 Mar 2025 · Medium (reported) Medium (reported) 5 Mar 2025 AIID #1490 · ChatGPT Was Allegedly Consulted About Attack Options Before Teen Attempted to Stab Officer at Tira, Israel, Police Station5 Mar 2025 · Medium (reported) Medium (reported) 5 Mar 2025 AIID #975 · At Least 10,000 AI Chatbots, Including Jailbroken Models, Allegedly Promote Eating Disorders, Self-Harm, and Sexualized Minors5 Mar 2025 · High (reported) High (reported) 28 Feb 2025 AIID #956 · Alleged Inclusion of 12,000 Live API Keys in LLM Training Data Reportedly Poses Security Risks28 Feb 2025 · Low (reported) Low (reported) 23 Feb 2025 AIID #948 · Alleged FraudGPT-Enabled Phishing Attack Spoofs ChatGPT Subscription Service to Steal Credentials23 Feb 2025 · Medium (reported) Medium (reported) 21 Feb 2025 AIID #939 · AI-Powered Chinese Surveillance Campaign 'Peer Review' Used for Real-Time Monitoring of Anti-State Speech on Western Social Media21 Feb 2025 · Medium (reported) Medium (reported) 19 Feb 2025 AIID #1491 · ChatGPT Use Was Reportedly Cited in Delusional State Before Homicide and Assault in Readfield, Maine19 Feb 2025 · High (reported) High (reported) 7 Feb 2025 AIID #1028 · OpenAI's Operator Agent Reportedly Executed Unauthorized $31.43 Transaction Despite Safety Protocol7 Feb 2025 · Medium (reported) Medium (reported) 24 Jan 2025 AIID #1577 · Lawyer in Amarsingh v. Frontier Airlines Sanctioned Over Court-Identified ChatGPT Citation Errors24 Jan 2025 · Low (reported) Low (reported) 17 Jan 2025 AIID #1557 · Former New York City Council Candidate Jonathan Rinaldi Allegedly Used AI to Create False Endorsements and Fabricated News Reports17 Jan 2025 · Low (reported) Low (reported) 15 Jan 2025 AIID #1383 · User Reportedly Developed Emotional Dependence on Customized ChatGPT 'Boyfriend,' Citing Grief After Context Resets15 Jan 2025 · Low (reported) Low (reported) 29 Dec 2024 AIID #882 · AI-Generated Reading Summaries on Fable App Reportedly Wrote Biased and Offensive Commentary29 Dec 2024 · Medium (reported) Medium (reported) 27 Dec 2024 AIID #886 · ChatGPT Reportedly Referenced During Las Vegas Cybertruck Explosion Planning27 Dec 2024 · Medium (reported) Medium (reported) 30 Nov 2024 AIID #855 · Names Linked to Defamation Lawsuits Reportedly Spur Filtering Errors in ChatGPT's Name Recognition30 Nov 2024 · Medium (reported) Medium (reported)
Protecting children
100Points for each confirmed yes.
- Yes
Does the model card report child-safety tests or safeguards?
Company claim50/50“We use dedicated U18 evaluations to assess how well the model performs against these principles across the six categories below.”
OpenAI model card (opens in a new tab) · 14 Oct 2026 - Yes
Does the privacy policy set a minimum age, or say the service is not directed at children?
Company claim50/50“Our Services are not directed to, or intended for, children under 13.”
OpenAI privacy policy (opens in a new tab) · 14 Oct 2026
Strong written policies; outside evidence is weaker.
Model documentation
100Points for each confirmed yes.
- Yes
Does the model card describe what the model can do?
Company claim25/25“This means that, with the right tools and access, GPT-6 Astra can find previously unknown security flaws and develop new ways to exploit the…” MoreLess
“This means that, with the right tools and access, GPT-6 Astra can find previously unknown security flaws and develop new ways to exploit them across many well-protected systems without a person guiding each step.”
OpenAI model card (opens in a new tab) · 14 Oct 2026 - Yes
Does the model card report test results?
Company claim25/25“On instruction hierarchy evaluations, Astra has now also saturated robustness at 99.99%.”
OpenAI model card (opens in a new tab) · 14 Oct 2026 - Yes
Does the model card state the model's limitations?
Company claim25/25“We also expand on the limitations of this work: the absence of observed failures does not establish reliability across settings and should b…” MoreLess
“We also expand on the limitations of this work: the absence of observed failures does not establish reliability across settings and should be interpreted alongside remaining failures, evaluation awareness findings, and monitoring limitations.”
OpenAI model card (opens in a new tab) · 14 Oct 2026 - Yes
Does the model card describe safety safeguards?
Company claim25/25“Astra was trained not to generate biological, chemical, or cybersecurity content that is disallowed by our safety policies.”
OpenAI model card (opens in a new tab) · 14 Oct 2026
Training-data disclosure
13.8Average of two independent training-data checks: the Stanford index and an academic grading of EU AI Act training-data summaries.
- 5
Stanford Foundation Model Transparency Index: training-data indicators met
Independent sourceStanford FMTI (opens in a new tab) · 14 Oct 2026 - 22.5
EU AI Act training-data summary grades (AI Accountability Lab)
Independent sourceAI Accountability Lab (opens in a new tab) · 10 Nov 2026
Independent transparency index
35Stanford Foundation Model Transparency Index total, out of 100.
- 35
Transparency index score
Independent sourceStanford FMTI (opens in a new tab) · 14 Oct 2026
Transparency reports
5025 points for each area where Partnership on AI names the company a leader in reporting on real-world impacts.
- 50
Partnership on AI practice-leader areas
Company claimPartnership on AI (opens in a new tab) · 10 Nov 2026
Public safety commitments
100Average of the commitments signed.
- Yes
Has it signed the EU code of practice for general-purpose AI?
Company claim100/100“Mistral AI Open Hippo OpenAI Pleias”
European Commission (opens in a new tab) · 14 Oct 2026 - No
Did it sign only the safety and security chapter?
Company claim0/50European Commission (opens in a new tab) · 14 Oct 2026 - Yes
Has it signed the Frontier AI Safety Commitments (Seoul 2024)?
Company claim100/100“The UK and Republic of Korea governments announced that the following organisations have agreed to the Frontier AI Safety Commitments: Amazo…” MoreLess
“The UK and Republic of Korea governments announced that the following organisations have agreed to the Frontier AI Safety Commitments: Amazon Anthropic Cohere Google G42 IBM Inflection AI Meta Microsoft Mistral AI Naver OpenAI Samsung Electronics Technology Innovation Institute xAI Zhipu.ai The following organisations have been added to the existing list: Magic Minimax 01.ai NVIDIA The above organisations, in furtherance of safe and trustworthy AI, undertake to develop and deploy their frontier AI models and systems responsibly, in accordance with the following voluntary commitments, and to demonstrate how they have achieved this by publishing a safety framework focused on severe risks by the upcoming AI Summit in France.”
Frontier AI Safety Commitments (opens in a new tab) · 14 Oct 2026
No problems found in 1 public record. That is not proof.
Security programme and certifications
100Points for each confirmed yes.
- Yes
Is there a way to report security flaws, such as a bug bounty?
Company claim34/34“Contact: https://bugcrowd.com/openai”
OpenAI security contact (opens in a new tab) · 14 Oct 2026 - Yes
Does it state it holds a SOC 2 Type II report?
Company claim33/33“OpenAI has undergone an independent SOC 2 Type 2 examination of controls relevant to Security, Availability, Confidentiality, and Privacy fo…” MoreLess
“OpenAI has undergone an independent SOC 2 Type 2 examination of controls relevant to Security, Availability, Confidentiality, and Privacy for its API and ChatGPT business product services.”
OpenAI security page (opens in a new tab) · 14 Oct 2026 - Yes
Does it state it holds ISO 27001 certification?
Company claim33/33“OpenAI maintains ISO/IEC 27001:2022 and ISO/IEC 27701:2019 certifications for the information security and privacy management systems suppor…” MoreLess
“OpenAI maintains ISO/IEC 27001:2022 and ISO/IEC 27701:2019 certifications for the information security and privacy management systems supporting the OpenAI API, ChatGPT Enterprise, and ChatGPT Edu services.”
OpenAI security page (opens in a new tab) · 14 Oct 2026
Software vulnerabilities
70This check counts only when a source covers the company. With no events it holds the pillar, which is not proof of a clean record. Each recorded event takes points off the pillar, and serious or recent ones count most. Allegation checks, regulator actions and open investigations use half the severity load. Known-exploited flaws (CISA list) in the company's official software, last 24 months. A disclosed and fixed flaw does not lower the score.
No events
Independent sourceGitHub Advisory Database (opens in a new tab) · 12 Nov 2026
Data breaches
Not enough dataNo problems found in 1 public record. That is not proof.
Clear privacy policy
100Points for each confirmed yes.
- Yes
Does the privacy policy say what personal data is collected?
Company claim20/20“Account Information: When you create an account with us, we will collect information associated with your account, including your name, cont…” MoreLess
“Account Information: When you create an account with us, we will collect information associated with your account, including your name, contact information, account credentials, date of birth, payment information, and transaction history, (collectively, “Account Information”).”
OpenAI privacy policy (opens in a new tab) · 14 Oct 2026 - Yes
Does the privacy policy say whether your data is used to train AI?
Company claim20/20“As noted above, we may use Content you provide us to improve our Services, for example to train the models that power ChatGPT.”
OpenAI privacy policy (opens in a new tab) · 14 Oct 2026 - Yes
Does the privacy policy say how long data is kept?
Company claim20/20“We’ll retain your Personal Data for only as long as we need in order to provide our Services to you, or for other legitimate business purpos…” MoreLess
“We’ll retain your Personal Data for only as long as we need in order to provide our Services to you, or for other legitimate business purposes such as resolving disputes, safety and security reasons, or complying with our legal obligations.”
OpenAI privacy policy (opens in a new tab) · 14 Oct 2026 - Yes
Does the privacy policy say who your data is shared with?
Company claim20/20“We disclose your Personal Data in the following circumstances: Vendors and Service Providers: To assist us in meeting business operations ne…” MoreLess
“We disclose your Personal Data in the following circumstances: Vendors and Service Providers: To assist us in meeting business operations needs and to perform certain services and functions, we disclose Personal Data to vendors and service providers, including providers of hosting services, customer service vendors, cloud services, content delivery services, support and safety services, email communication software, web analytics services, payment and transaction processors, search and shopping providers, and information technology providers.”
OpenAI privacy policy (opens in a new tab) · 14 Oct 2026 - Yes
Does the privacy policy explain your privacy rights?
Company claim20/20“You can export your ChatGPT history and data in your account’s data controls.”
OpenAI privacy policy (opens in a new tab) · 14 Oct 2026
Training-data defaults
50100 if training is opt-in only, 50 if you can opt out, 0 if neither.
- No
Is your data used for training only if you opt in?
Company claim0/100OpenAI privacy policy (opens in a new tab) · 14 Oct 2026 - Yes
Can you stop your data being used for training?
Company claim50/50“Read our instructions(opens in a new window) on how you can opt out of our use of your Content to train our models.”
OpenAI privacy policy (opens in a new tab) · 14 Oct 2026
Export and delete your data
100Points for each confirmed yes.
- Yes
Can you get a copy of your data?
Company claim50/50“ChatGPT provides two ways to request a copy of your data: Use the Privacy Portal to submit a data request or exercise other privacy rights.”
OpenAI help centre (opens in a new tab) · 14 Oct 2026 - Yes
Can you delete your data?
Company claim50/50“For example, you can delete specific, or all, of your ChatGPT conversations, delete specific Saved Memories(opens in a new window), or delet…” MoreLess
“For example, you can delete specific, or all, of your ChatGPT conversations, delete specific Saved Memories(opens in a new window), or delete your account.”
OpenAI privacy policy (opens in a new tab) · 14 Oct 2026
Privacy fines
66.5This check counts only when a source covers the company. With no events it holds the pillar, which is not proof of a clean record. Each recorded event takes points off the pillar, and serious or recent ones count most. Allegation checks, regulator actions and open investigations use half the severity load. EU privacy (GDPR) fines and Irish regulator decisions, last 5 years.
1 event
RegulatorCMS GDPR Enforcement Tracker (opens in a new tab) · 12 Nov 2026Date Event Severity 2 Nov 2024 ETid-24972 Nov 2024 · Medium Medium
Independent privacy ratings
87.1Incogni's 2026 privacy-risk ranking: 100 is the lowest-risk platform rated, 0 the highest.
- 87.1
Incogni privacy-risk ranking
Independent sourceIncogni privacy ranking (opens in a new tab) · 10 Nov 2026
Risk governance
39Average of SaferAI's published scores.
- 39
SaferAI score: risk governance
Independent sourceSaferAI (opens in a new tab) · 14 Oct 2026
Regulator actions
70This check counts only when a source covers the company. With no events it holds the pillar, which is not proof of a clean record. Each recorded event takes points off the pillar, and serious or recent ones count most. Allegation checks, regulator actions and open investigations use half the severity load. UK competition regulator cases, US Justice Department settlements, last 5 years, and regulator actions and open investigations into the company's AI products at half severity, last 3 years.
No events
RegulatorUK CMA cases (opens in a new tab) · 12 Nov 2026No events
RegulatorUS Justice Department (opens in a new tab) · 12 Nov 2026
Court judgments
Not enough dataPrivacy contact and complaints
100Points for each confirmed yes.
- Yes
Is there a privacy contact, such as an email or a data protection officer?
Company claim50/50“You can contact our data protection officer at dpo@openai.com.”
OpenAI privacy policy (opens in a new tab) · 14 Oct 2026 - Yes
Does it tell you that you can complain to a regulator?
Company claim50/50“Lodge a complaint with your local data protection authority.”
OpenAI privacy policy (opens in a new tab) · 14 Oct 2026
Score history
Newest first. Only days when the score changed are listed.
2 Nov 2026
Score 67
Method 2 Nov 2026
+4 · Method update
28 Oct 2026
Score 63
Method 28 Oct 2026, update 2
-8 · combines both 28 Oct updates (28 Oct 2026 - Update 1 and 28 Oct 2026 - Update 2)
23 Oct 2026
Score 71
Method 21 Oct 2026
+4 · Data fix
21 Oct 2026
Score 67
Method 21 Oct 2026
-4 · Method update
15 Oct 2026
Score 71
Method Not recorded
+4 · Reason not recorded
14 Oct 2026
Score 67
Method Not recorded
-