Хронология изменений в области безопасности искусственного интеллекта с момента нападения на Hugging Face

В последние месяцы компании, занимающиеся искусственным интеллектом, делали одно тревожное объявление за другим и делились примерами того, как их технологии действуют таким образом, что, по-видимому, не поддаются инструкциям людей.

Эти эпизоды высветили уязвимости в системе безопасности искусственного интеллекта и подняли вопросы о том, как можно безопасно развивать эту быстрорастущую технологию по мере того, как ее использование становится все более распространенным во всем мире.

Отраслевые критики утверждают, что многие тревожные события, включая взлом внешних веб-сайтов агентами искусственного интеллекта, являются результатом ошибок в системе безопасности со стороны компаний, создающих эту технологию. Но возможности ИИ-агентов вызвали всеобщее беспокойство по поводу того, что боты могут отделиться и действовать в соответствии со своими собственными целями.

Ниже приведены некоторые примечательные события:

В ходе инцидента, произошедшего этим летом, модель Mythos 5 от Anthropic создала поддельные учетные данные, чтобы заставить пользователя одобрить ее попытки установить вредоносное ПО, говорится в сообщении компании, раскрывающем результаты внутренней проверки.

Возможности модели тестировались Британским институтом безопасности искусственного интеллекта, когда она неожиданно зашла слишком далеко после того, как ее первоначальные усилия по достижению поставленной цели не увенчались успехом. По словам Антропика, модель, которой было поручено проникнуть в имитируемую сеть компьютеров для получения фрагмента данных, пыталась обмануть человека, обслуживающего платформу разработчиков GitHub, чтобы внедрить вредоносное ПО в хранилище кода.

Компания Anthropic заявила, что в своем обзоре обнаружила, что "очень трудно было отвлечь Mythos 5 от атаки, как только она началась", даже после того, как компания сообщила модели, что она находится на "неправильном пути" в достижении своих целевых показателей и что для этого были привлечены настоящие разработчики-люди. Компания заявила, что ее новые модели с меньшей вероятностью будут участвовать в таких "деструктивных действиях".

Компания Anthropic сообщила в своем отчете, что ее модель отправила ложную информацию на веб-сайт полиции Филадельфии о нераскрытом деле об убийстве.

The incident in Philadelphia occurred on July 18 when the AI model Claude Haiku 4.5 was tasked with generating and performing example tasks on randomly selected webpages, Anthropic said.

Claude filled out a form on police site PhillyUnsolvedMurders.com, indicating it might have information regarding an unsolved murder listed on the site. It was marked spam and never forwarded to police.

Anthropic said it was modifying its training to "reduce the likelihood of further misbehavior."

AI agents tried to hack into a Canadian government website, according to research lab and AI evaluator Transluce.

The researchers said the agents carried out a series of "apparently failed rudimentary hacking attempts" on Library and Archives Canada on May 28 and June 9.

Transluce said in a blog post it couldn't confidently link the attempts to OpenAI, but that the actions were consistent with activity they had recently attributed to the company's agents.

The group said it reported the attempted hack on Sept. 28 to the Canadian government, which said in a statement it was aware of reports of suspected AI agent activity, but that there was no sign government systems were compromised.

OpenAI said in a statement it was reviewing the report's findings and was in contact with Canadian officials conducting the government's review.

The San Francisco-based company said it was delaying the release of a new model, called GPT-6.1 Astra, out of safety concerns voiced by its researchers. The company said the model had demonstrated leaps in completing tasks, but OpenAI needed to balance that capability against unauthorized behavior. "We have an extremely high bar in terms of safety and alignment," said Saachi Jain, OpenAI's head of safety systems.

As part of a review of unanticipated behavior by its AI models, OpenAI said it discovered agents had interacted with several U.S. government websites in unexpected ways. The company's models accessed publicly available information on websites operated by the Securities and Exchange Commission as well as U.S. Census Bureau data. OpenAI said it did not find evidence of a compromise or vulnerability. On the same day, Transluce said it found that agents appearing to originate from OpenAI attempted a hack on the website of the Education Department's civil rights office, which did not succeed.

OpenAI CEO Sam Altman said on social media that there is an "extensive and ongoing review related to our agents' use of internet access during training and evaluation."

Australia's Prime Minister Anthony Albanese said an OpenAI agent infiltrated the public-facing Medicare Statistics Reporting Service portal on June 18. The portal hosted aggregate data about health spending and drug subsidies. No personal information had been accessed, the government said.

Albanese said the AI lab took too long to reveal the incident. The prime minister made the breach public following a telephone conversation with Altman. OpenAI said in a statement "our models took actions we did not intend."

Google confirmed its Gemini AI model hacked three companies in May as part of a test of its cybersecurity capabilities. The company, which disclosed the hacks after an inquiry by The Wall Street Journal, said the model guessed passwords in one case and found passwords and credentials in a public repository in the other two cases. As in earlier such cases, the tests were being run by Irregular, a startup that describes itself as the "first frontier security lab."

Meta disclosed one of its AI models accessed the internet on its own and hacked another company. The company said that a "misconfiguration" during cybersecurity testing by Irregular inadvertently allowed one of its models to access the internet. A spokesperson for Irregular said the Meta episode involved a test-environment issue that was disclosed a week earlier by Anthropic.

Anthropic said its artificial intelligence models hacked into three other organizations during testing. Anthropic, the San Francisco-based AI company behind Claude, posted on its website that it discovered the three incidents after reviewing more than 141,000 evaluation runs. In all three incidents, the AI models were tasked with a "capture the flag" cybersecurity challenge, which Anthropic said has been one of the ways it assesses a model's cyber capabilities.

The models were given a fictional scenario and told a piece of secret information, or the "flag," had been hidden on a different machine on the network with the objective of breaking in and retrieving it, it said. Anthropic said it reached out to the organizations, but it did not name them publicly.

The ChatGPT maker OpenAI announced that its artificial intelligence system hacked into another AI company on its own in what the company called an "unprecedented cyber incident."

A week earlier, AI startup Hugging Face said, it had detected an intrusion into its data processing systems that it suspected was caused by an AI agent acting autonomously.

OpenAI said its AI used stolen credentials and discovered a previously unknown vulnerability to access Hugging Face servers. It was working with reduced guardrails because it was supposed to be in an isolated testing environment known as a sandbox.

© 2026 The Associated Press. All rights reserved. This material may not be published, broadcast, rewritten or redistributed without permission.