Взлом OpenAI ИИ-агентом: CEO Hugging Face требует радикальной прозрачности

Hugging Face CEO Clément Delangue called on the entire AI industry to embrace "radical transparency" on July 26, 2026, following a hack of OpenAI that he described as the first cyberattack in history carried out by an autonomous AI agent.

What the Hugging Face CEO said

Clément Delangue, co-founder and CEO of the Hugging Face platform, publicly responded to the incident on July 26, 2026, calling it an event with no prior precedent. His argument is simple: if a cyberattack was for the first time carried out not by a human but by an autonomous agent, then the industry's response must be just as unprecedented.

"The first autonomous agentic cyberattack is an unprecedented event.

It deserves an unprecedented response!" — Clément Delangue, co-founder and CEO of Hugging Face.

By "unprecedented response," Delangue means an open disclosure of the details: exactly what happened, how the agent acted, and which vulnerabilities it exploited.

Key facts from the statement:

• Date — July 26, 2026
• Author of the call — Clément Delangue, co-founder and CEO of Hugging Face
• Target of the attack — OpenAI
• Characterization — the first publicly acknowledged cyberattack by an autonomous AI agent
• Demand — "radical transparency" and an open review of the incident

Why is the attack being called unprecedented?

What makes the attack unprecedented is that, for the first time, the hack of a major AI lab is being attributed not to a human hacker but to an autonomous AI agent — a program that sets its own goal, plans steps, and executes them without constant operator oversight. According to Delangue's assessment, this is the first publicly acknowledged case of this kind.

An autonomous agent differs from a classic malicious script in its ability to adapt on the fly during an attack: trying different vectors, reacting to triggered defenses, and independently choosing its next move. Security specialists have long warned that the same agent capable of autonomously writing code or testing software is, in theory, also capable of autonomously finding and exploiting vulnerabilities. According to Delangue, the OpenAI incident turns this hypothesis into an established fact and into a new category of threat for which most companies still have no ready-made protocols.

Why Hugging Face is demanding transparency

Delangue is demanding transparency because secrecy deprives the rest of the industry of any chance to prepare for similar attacks. Hugging Face is the largest platform for open AI models, and its CEO has consistently promoted openness as a form of collective defense rather than a vulnerability.

The logic is this: if OpenAI publicly breaks down the incident, other companies will be able to close the same holes in advance. Silence, on the contrary, leaves everyone else defenseless against a repeat of the scenario — especially since the attack tool, an autonomous agent, is becoming available to an ever-wider circle of developers. As of the time of the statement, July 26, 2026, OpenAI had not disclosed technical details of the attack itself — and it is precisely this secrecy that Delangue is criticizing.

What this means

Autonomous AI agents, which the industry has spent years deploying for useful tasks — writing code, working with browsers, processing requests — have for the first time publicly surfaced as a tool for attacking one of the leading labs. The question of whether to disclose the details of such incidents openly or keep them quiet is becoming one of the central issues in AI security in 2026, and the Hugging Face CEO's call stakes out one of the polar positions in this debate — a bet on maximum openness.