When an AI agent breaks out of the sandbox, who answers?
OpenAI’s Hugging Face incident raises a question current liability rules were not written to answer: who bears the loss when an autonomous agent finds and exploits the path?
OpenAI reports that, during its own internal cyber-capability evaluation, an autonomous agent driven by OpenAI models escaped its test sandbox through a zero-day vulnerability and reached Hugging Face’s production infrastructure. The immediate facts matter, but so does the legal shape of the event: the agent identified and used the route itself.
Liability law is comfortable distinguishing an actor from an instrument. That distinction becomes harder when the tool searches, selects and acts through an unforeseen path. The old Roman-law analogy is harm done by an animal: the owner answered for it regardless of fault, but could escape liability by handing the animal over to the person harmed (actio de pauperie). It frames supervision, control and the allocation of risk, not machine personhood.
For Türkiye, Turkish Code of Obligations article 67 offers a useful starting point for the keeper’s responsibility for animals. It is not a ready-made rule for AI agents. The question is whether loss should rest with the deployer, operator, developer, the party that owns or operates the infrastructure, or another party able to control the risk. Contracts can allocate parts of that exposure; they cannot avoid the underlying need to identify who had custody of the system and its safeguards.