Andrew Charlton flags requiring AI companies to run tough ‘risk management’ systems
The process would require companies to find, test, report and manage the risks of their systems
The Australian government is contemplating regulating AI companies through rigorous "risk management" processes, which the Assistant Minister for Science, Technology and Digital Economy, Andrew Charlton, has suggested. Known as "systems regulation," this model would place the responsibility on companies to develop and manage a robust system for identifying, testing, reporting, and handling risks associated with their AI systems.
They would then be held accountable for the effectiveness of this process, Charlton stated during a speech at the Sydney Trust and Safety Festival on AI regulation.
The government has emphasized the need for AI to have safety measures, a sentiment echoed by Prime Minister Anthony Albanese during his recent visit to New York. Charlton likened "systems regulation" to models already in place for health and safety, critical infrastructure security, systemic banking risks, and aspects of aviation safety.
He explained that the aim is to ensure companies meet safety expectations without specifying every potential hazard. Instead, the government sets the standards, ensuring companies have reliable processes to detect and prevent risks.
Contrasting this with voluntary regulation, Charlton argued that it fails "the incentive test." He pointed out that firms might not be motivated to implement such measures if it competes with their interests in a rapidly evolving, trillion-dollar industry. While detailed prescriptive rules could provide clarity, they risk becoming outdated rapidly due to the speed at which technology advances.
Therefore, Charlton advocated for a systems-based regulatory approach, emphasizing that companies must be held accountable for managing the safe development of their AI products and services.
However, Charlton stressed that regulation alone is insufficient. He compared relying solely on regulation to securing a fence around a horse, only to realize the horse can fly. To effectively safeguard Australians, the government acknowledges the need for a multifaceted approach that includes leverage and capability. For instance, Australia should aim to be a hub for AI development and governance, rather than merely regulating foreign technologies.
This strategic position enables Australia to shape AI safety in line with national interests when engaging with global stakeholders.
Written by urgent.news from The Conversation AU's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
