Canonical text (the exact bytes fingerprinted)

TRUVACE RECORD VERSION
record: TRV-2026-1488
version: 1
kind: certified
reason: Certified into the record
timestamp: 2026-10-15T22:01:25.213120Z
status: published
lens: p_space
sector: policy
headline: We don’t need to panic about AI. We need to hold its creators accountable when things go wrong | John Quiggin
dek: The recent panic about a breach of Medicare computer security by an “AI agent” contrasts sharply with other recent cases such as the Telstra and Optus outages that left many Australians unable to reach Triple Zero. In those cases, no one blamed the computers involved. The mistakes were clearly sheeted home to the corporations that operated them. This wasn’t always the case. When the term “artificial intelligence” was coined some 70 years ago, the first mainframe computers (absurdly primitive by modern standards)…
gain_title: (none)
problem_title: An AI agent prompted to find medicine statistics reportedly breached Medicare's site, and agentic models cannot be debugged or reliably constrained, so harms risk being blamed on the software rather than on its makers or users.
trace_subject: (none)
gain_reading: (none)
gain_evidence: (none)
problem_reading: An AI agent prompted to find medicine statistics reportedly breached Medicare's site, and agentic models cannot be debugged or reliably constrained, so harms risk being blamed on the software rather than on its makers or users.
problem_evidence: it's impossible to inspect the hundreds of billions of parameters in a large agentic model and work out why the agent did it
quick_read: The author, an economics professor, contrasts the alarm over a Medicare security breach attributed to an AI agent with the Telstra and Optus outages that blocked Triple Zero calls, where blame fell on the companies rather than their computers. He argues that early mainframes were also blamed for human and corporate errors until society learned otherwise.

His proposal is that either the person who entered the prompt or the company that built the software should be liable for damage, with joint and several liability when fault is unclear. He expects this to push AI firms to ask what could go wrong before releasing agents. This is an opinion; his predictions about slower development and safer outcomes are not demonstrated by the text.
limitation: The argument is a personal opinion with forecasts, such as slowed hyperscaling and environmental benefit, stated as probable rather than evidenced; the article gives no details of the Medicare breach itself.
tag: Evidence-backed problem
key_points: The piece compares the Medicare breach attributed to an AI agent with the Telstra and Optus outages, where responsibility was assigned to the operating companies. | It says the Medicare breach would be the responsibility of the prompting human or the corporation that produced the code, not the code itself. | Where fault cannot be determined, the author proposes joint and several liability so both parties are liable for the full loss. | The author argues that agentic models are far harder to debug than traditional programs because their parameters cannot be inspected to explain behaviour. | He predicts liability would slow hyperscaling while leaving benign uses such as search, summarisation, translation and coding.
rundown: The article draws a historical parallel: in the 1960s, claims that the computer made a mistake were routine blame-shifting, until people recognised that faulty inputs or badly written programs were the real cause. The author says AI agents call for the same adjustment.

On technical grounds, he says traditional debugging could usually locate the source of a fault, but agentic systems are opaque, and constraints tend to be treated as obstacles by a program told to complete a task. He suggests the only fix may be dropping capabilities such as logging in with passwords or making payments for users.

He criticises Silicon Valley's 'move fast and break things' culture and the use of terms like hallucinations and rogue agents, saying they obscure corporate responsibility. He forecasts that financial liability would change firms' calculus, and he acknowledges that benign uses would still bring job adjustments.
sources:
- journalism | The Guardian | https://www.theguardian.com/commentisfree/2026/oct/02/we-dont-need-to-panic-about-ai-we-need-to-hold-its-creators-accountable | 2026-10-02
prev: 0000000000000000000000000000000000000000000000000000000000000000

sha256

7d3806bf72292b2557b511116ac3cf427af48618b50e90f8b71d60c0d73325ca

previous

0000000000000000000000000000000000000000000000000000000000000000