Opinion: AI Accountability Should Rest With Developers, Not the Technology Itself

A commentary responding to a recent security breach affecting Australia's Medicare computer system argues that responsibility for AI-related failures should rest with the companies and developers behind the technology, not with the AI systems themselves.

The piece contrasts public reaction to the Medicare breach, reportedly linked to an "AI agent," with reactions to earlier outages affecting Telstra and Optus that left many Australians unable to reach triple-zero emergency services. In those cases, the article notes, blame was clearly directed at the corporations operating the systems rather than at the technology involved.

The author argues that a similar standard should apply to AI-related incidents: when an AI system produces a harmful or unintended outcome, such as a prompt triggering a security breach, the responsibility lies with the organizations that built and deployed the system, not with the code itself.

The commentary situates the current debate over AI within a longer history of public anxiety about new computing technology, recalling that early mainframe computers decades ago were greeted with a mixture of awe and concern similar to that now directed at AI agents.

The piece calls for greater accountability from AI developers and the companies that deploy AI systems, suggesting that this approach, rather than broader panic about the technology itself, is the more productive response to incidents like the Medicare breach.