{"check":null,"uid":"92d75ac11a6eaddd","title":"Defense AI Reliability and Reporting Act","title_generated":false,"country":"США","organ":"Конгресс США","kind":"law","kind_name":"Законодательство","lang":"en","date":"2026-08-31","summary":"Минобороны США создаёт единую программу учёта и устранения инцидентов и уязвимостей систем искусственного интеллекта — от разработки и испытаний до закупки и эксплуатации.\nОтветственный чиновник получает обязательные срочные сообщения о каждом инциденте и уязвимости; случаи делятся на общеведомственные, программные и локальные. Для первых двух категорий требуется план устранения и проверка мер до продолжения эксплуатации системы. Военнослужащие, гражданские сотрудники и подрядчики могут сообщать о проблемах без риска дисциплинарных мер.\nС 2027 по 2031 год министерство ежегодно отчитывается перед профильными комитетами Конгресса, включая случаи гибели или травм военнослужащих.","snippet":"","topics":["Искусственный интеллект"],"status":"ok","error":"","text_len":7589,"versions":1,"url":"https://www.congress.gov/bill/119th-congress/house-bill/10189","first_seen":"2026-09-02","last_checked":"2026-09-17 01:57","relevance":"hit","score":13,"query":"","source_key":"congress_us","verdict":{"relevance":"hit","score":13,"topics":["Искусственный интеллект"],"need_body":3,"authorities":[],"evidence":[{"topic":"Искусственный интеллект","term":"искусственн","weak":false,"pos":132,"ctx":"инобороны сша создаёт единую программу учёта и устранения инцидентов и уязвимостей систем искусственного интеллекта — от разработки и испытаний до закупки и эксплуатации. ответственный чиновник","zone":"название","weight":3},{"topic":"Искусственный интеллект","term":"artificial intelligence","weak":false,"pos":267,"ctx":"to amend title 10, united states code, to direct the secretary of defense to establish an artificial intelligence incident and vulnerability reporting program, and for other purposes.  __________________","zone":"текст","weight":1},{"topic":"Искусственный интеллект","term":"artificial intelligence","weak":false,"pos":798,"ctx":"to amend title 10, united states code, to direct the secretary of defense to establish an artificial intelligence incident and vulnerability reporting program, and for other purposes.  be it enacted by t","zone":"текст","weight":1},{"topic":"Искусственный интеллект","term":"artificial intelligence","weak":false,"pos":1317,"ctx":"ode, is amended by inserting after section 2224a the following new section: ``sec. 2224b. artificial intelligence incident and vulnerability reporting program ``(a) in general.--the secretary of defense","zone":"текст","weight":1},{"topic":"Искусственный интеллект","term":"artificial intelligence","weak":false,"pos":1672,"ctx":"erabilities arising from the development, testing, procurement, fielding, or operation of artificial intelligence systems within the department of defense. ``(b) purpose.--the purpose of the program esta","zone":"текст","weight":1},{"topic":"Искусственный интеллект","term":"artificial intelligence","weak":false,"pos":1919,"ctx":"`(1) identify recurring risks, failure modes, vulnerabilities, and systemic weaknesses in artificial intelligence systems, including risks or failure modes arising from human-machine teaming; ``(2) suppo","zone":"текст","weight":1},{"topic":"Искусственный интеллект","term":"artificial intelligence","weak":false,"pos":6226,"ctx":"ns for changes to testing, procurement, cybersecurity, or deployment policies relating to artificial intelligence systems. ``(2) each report under this subsection shall be submitted in unclassified form","zone":"текст","weight":1}],"dropped":[{"topic":"Кибербезопасность","term":"cybersecurity","weak":false,"pos":2107,"ctx":"ng; ``(2) support mitigation of significant risks; and ``(3) inform testing, procurement, cybersecurity, and deployment decisions to improve the safety, security, reliability, and operational e","why":"одиночное упоминание (нужно 3)"},{"topic":"Кибербезопасность","term":"cybersecurity","weak":false,"pos":6176,"ctx":"rective actions taken; and ``(d) any recommendations for changes to testing, procurement, cybersecurity, or deployment policies relating to artificial intelligence systems. ``(2) each report un","why":"одиночное упоминание (нужно 3)"}]},"last_changed":"2026-09-02","meta":{"congress":"119","billType":"HR","number":"10189","policyArea":"","subjects":[],"latestAction":"2026-08-31 Referred to the House Committee on Armed Services."},"source_url":"https://www.congress.gov/bill/119th-congress/house-bill/10189","text":"FULL TEXT\n[Congressional Bills 119th Congress]\n[From the U.S. Government Publishing Office]\n[H.R. 10189 Introduced in House (IH)]\n\n<DOC>\n\n119th CONGRESS\n2d Session\nH. R. 10189\n\nTo amend title 10, United States Code, to direct the Secretary of\nDefense to establish an artificial intelligence incident and\nvulnerability reporting program, and for other purposes.\n\n_______________________________________________________________________\n\nIN THE HOUSE OF REPRESENTATIVES\n\nAugust 31, 2026\n\nMs. Jacobs (for herself, Mr. Moran, and Mr. Whitesides) introduced the\nfollowing bill; which was referred to the Committee on Armed Services\n\n_______________________________________________________________________\n\nA BILL\n\nTo amend title 10, United States Code, to direct the Secretary of\nDefense to establish an artificial intelligence incident and\nvulnerability reporting program, and for other purposes.\n\nBe it enacted by the Senate and House of Representatives of the\nUnited States of America in Congress assembled,\n\nSECTION 1. SHORT TITLE.\n\nThis Act may be cited as the ``Defense AI Reliability and Reporting\nAct''.\n\nSEC. 2. DEPARTMENT OF DEFENSE AI INCIDENT AND VULNERABILITY REPORTING\nPROGRAM.\n\nChapter 131 of title 10, United States Code, is amended by\ninserting after section 2224a the following new section:\n``Sec. 2224b. Artificial intelligence incident and vulnerability\nreporting program\n``(a) In General.--The Secretary of Defense shall establish a\ncentralized Department-wide program for the reporting, tracking,\nanalysis, and remediation of covered AI incidents and covered AI\nvulnerabilities arising from the development, testing, procurement,\nfielding, or operation of artificial intelligence systems within the\nDepartment of Defense.\n``(b) Purpose.--The purpose of the program established under\nsubsection (a) shall be to--\n``(1) identify recurring risks, failure modes,\nvulnerabilities, and systemic weaknesses in artificial\nintelligence systems, including risks or failure modes arising\nfrom human-machine teaming;\n``(2) support mitigation of significant risks; and\n``(3) inform testing, procurement, cybersecurity, and\ndeployment decisions to improve the safety, security,\nreliability, and operational effectiveness of such systems.\n``(c) Requirements for Program.--The program shall--\n``(1) be designed using practices drawn from established\nsafety incident reporting programs, vulnerability disclosure\nprograms, and programs to identify and develop lessons learned;\n``(2) emphasize non-punitive reporting, protection of\nsensitive and proprietary information, and dissemination of\nlessons learned, as appropriate; and\n``(3) include a mechanism to enable timely access to and\nsharing of relevant logs, system data, and model information as\nnecessary to support analysis and response.\n``(d) Designation of Official.--The Secretary shall designate an\nappropriate official for the reporting, tracking, analysis, and\nremediation of covered AI incidents and covered AI vulnerabilities\nunder this section. The Secretary, acting through such official, shall\nreceive and standardize reports, conduct trend analysis, identify\nrecurring risks and failure modes, and issue guidance, alerts, and\nrecommendations, as appropriate.\n``(e) Reporting and Categorization.--(1) The Secretary shall\nrequire prompt reporting to the official designated under subsection\n(d) of--\n``(A) any covered AI incident; and\n``(B) any covered AI vulnerability.\n``(2) The Secretary, acting through the official, shall categorize\neach incident or vulnerability reported to the official according to\nwhether the incident or vulnerability requires--\n``(A) a Department-wide response;\n``(B) a response at the program level; or\n``(C) a response at a local level.\n``(f) Department-Wide and Program-Level Matters.--(1) In the case\nof any incident or vulnerability categorized under subsection (e)(2)(A)\nor (B), the Secretary, acting through the official designated under\nsubsection (d), shall coordinate any responses that the Secretary\nconsiders appropriate, such as remediation, retesting, mitigation\nmeasures, or deployment restrictions.\n``(2) In addition, in the case of any incident or vulnerability\ndescribed in subsection (e)(2)(A), the Secretary, acting through the\nofficial, shall require--\n``(A) a documented corrective action plan; and\n``(B) validation that the mitigation measures, if any, in\nsuch plan have been implemented before continued operational\nuse.\n``(g) Protection of Reports.--(1) The Secretary shall establish a\nprotected disclosure process, informed by established vulnerability\ndisclosure practices, through which members of the Armed Forces,\ncivilian employees, contractors, and subcontractors at any tier may\nreport covered AI incidents and covered AI vulnerabilities in good\nfaith.\n``(2) The Secretary shall ensure that a person making a report in\ngood faith under paragraph (1) is not, on the basis of that report\nalone, subject to adverse contract action, subject to adverse personnel\naction, or otherwise retaliated against by the Department.\n``(h) Protection of Information.--The Secretary shall establish\nprocedures to protect sensitive, proprietary, and classified\ninformation submitted through the protected disclosure process under\nsubsection (g).\n``(i) Annual Report.--(1) In each of years 2027 through 2031, the\nSecretary shall submit to the congressional defense committees an\nannual report on the program. The report shall include--\n``(A) the number of reports made of incidents and\nvulnerabilities and the categorizations of such reports;\n``(B) a summary of significant trends, recurring risks,\nsystemic issues, and corrective actions taken in response;\n``(C) in the case of any covered AI incident resulting in\nthe loss of life of, or in bodily harm to, a member of the\nArmy, Navy, Marine Corps, Air Force, or Space Force--\n``(i) a description of the incident, including the\nsystem or systems involved and the operational context;\n``(ii) the date and time the incident occurred;\n``(iii) an assessment of the cause and operational\nconsequence of the incident; and\n``(iv) any corrective actions taken; and\n``(D) any recommendations for changes to testing,\nprocurement, cybersecurity, or deployment policies relating to\nartificial intelligence systems.\n``(2) Each report under this subsection shall be submitted in\nunclassified form but may include a classified annex.\n``(j) Definitions.--In this section:\n``(1) The term `artificial intelligence' has the meaning\ngiven such term in section 5002 of the National Artificial\nIntelligence Initiative Act of 2020 (15 U.S.C. 9401).\n``(2) The term `covered AI incident' means an event in\nwhich an artificial intelligence system--\n``(A) causes unintended operational, safety, or\nsecurity harm;\n``(B) operates outside authorized parameters or\napproved safety, legal, or mission guardrails;\n``(C) materially degrades mission performance or\nreliability in a real-world or operationally\nrepresentative environment;\n``(D) fails to respond to an operator disengage\ncommand;\n``(E) operates in a manner that, under reasonably\nforeseeable circumstances, could have resulted in\nsignificant unintended operational, safety, or security\nharm; or\n``(F) operates in a manner that raises concerns\nregarding system control and autonomy.\n``(3) The term `covered AI vulnerability' means an\nexploitable weakness, vulnerability, or systemic issue in an\nartificial intelligence system or related component that could\nmaterially affect mission performance, compromise system\nintegrity, create safety risk, or result in unauthorized or\nunintended behavior.''.\n<all>","changes":[],"passport":{"data":{"act":{"jurisdiction":"США","title_official":"Defense AI Reliability and Reporting Act","title_short":"Закон о надежности и отчетности в сфере оборонного искусственного интеллекта","level":"закон","date_adopted":"2026-08-31","date_in_force":"","date_version":"","phased":"","status":"законопроект: на рассмотрении","sunset":"","regulator":"Министерство обороны США","related":""},"goal":{"problem":"необходимость повышения безопасности и надёжности систем искусственного интеллекта Министерства обороны США путём создания централизованной программы учёта и устранения инцидентов и уязвимостей таких систем.","goal":"обеспечение своевременной идентификации рисков, слабых мест и недостатков систем искусственного интеллекта, поддержка их минимизации, улучшение решений по тестированию, закупке, кибербезопасности и внедрению для повышения безопасности, защищённости, надёжности и эффективности этих систем.","targets":"","scope":"разработка, тестирование, приобретение, внедрение и эксплуатация систем искусственного интеллекта Министерством обороны США.","exclusions":""},"subjects_note":{"protected":"военнослужащие"},"subjects":[{"role":"поставщик","who":"разработчики, производители и интеграторы систем искусственного интеллекта","criteria":"участие в разработке, производстве или интеграции систем искусственного интеллекта для нужд Министерства обороны США","count":"нет данных"},{"role":"госорган","who":"Министерство обороны США","criteria":"","count":"нет данных"}],"norms":[{"address":"Sec. 2224b(e)(1)","addressee":"поставщик","essence":"Поставщики обязаны незамедлительно сообщать ответственному должностному лицу обо всех инцидентах и уязвимостях систем искусственного интеллекта","type":"обязанность","mechanism":"информирование","cost_channel":"административные","cost_kind":"регулярные","trigger":"при инциденте","sanction":"","refs":"","form":"цифровая","in_force":"","ru_analog":"требует проверки"},{"address":"Sec. 2224b(f)(2)","addressee":"поставщик","essence":"В случае инцидента или уязвимости, требующих ответственного реагирования на уровне ведомства или программы, поставщики должны разработать документированный план корректирующих действий и подтвердить выполнение предложенных мер до возобновления использования системы","type":"обязанность","mechanism":"ограничение модели","cost_channel":"содержательные","cost_kind":"по событию","trigger":"при инциденте","sanction":"","refs":"","form":"цифровая","in_force":"","ru_analog":"требует проверки"},{"address":"Sec. 2224b(g)(1)","addressee":"поставщик","essence":"Разработан защищённый процесс раскрытия информации, позволяющий сотрудникам и подрядчикам сообщать о происшествиях и недостатках систем искусственного интеллекта добросовестно и без страха наказания","type":"право","mechanism":"защита прав","cost_channel":"нет прямых издержек","cost_kind":"","trigger":"постоянно","sanction":"","refs":"","form":"цифровая","in_force":"","ru_analog":"требует проверки"},{"address":"Sec. 2224b(i)(1)","addressee":"госорган","essence":"Министерство обороны обязано ежегодно предоставлять отчётность Конгрессу о программе регистрации и анализа инцидентов и уязвимостей систем искусственного интеллекта за период с 2027 по 2031 годы","type":"обязанность","mechanism":"информирование","cost_channel":"административный","cost_kind":"регулярные","trigger":"ежегодно","sanction":"","refs":"","form":"цифровая","in_force":"","ru_analog":"требует проверки"}]},"made_by":"GigaChat-2-Max","made_at":"2026-09-10 11:44:11","edited_at":null,"edited_by":null}}