Australia missing early warning system for AI threats: Intelligence chief

Canberra, Sep 15 (IANS) The head of Australia’s cyber intelligence agency has warned that the country needs an early warning system for threats posed by AI.

Abigail Bradshaw, director-general of the Australian Signals Directorate (ASD), told a summit in Canberra on Monday that AI attacks could exploit outdated technology being used by governments and businesses.

She called for organizations to embrace AI as a defensive tool and said that intelligence and security agencies must work with AI companies to identify early risks, Xinhua news agency reported.

It comes as Australian politicians and AI company executives respond to a growing international chorus of warnings the race to develop AI could outstrip humanity’s ability to control it.

“What we are missing at the moment, which we have in a cyber context but we don’t yet have it formalized in an AI context, is an early warning system,” Bradshaw said.

The summit in Canberra was held days after Anthropic CEO Dario Amodei on Saturday called for the industry to slow down and for third-party evaluators to verify models’ adherence to safety measures.

OpenAI CEO Sam Altman and xAI’s Elon Musk both said they agreed with Amodei.

In response, Australian Assistant Minister for Science, Technology and the Digital Economy Andrew Charlton said on Monday that he does not want technology companies or the United States to be solely responsible for setting AI guardrails.

“I don’t think we should care very much about what individual chief executives say or what individual executives claim they want to happen,” he said.

“I am not in favor of leaving it to the companies and I’m not in favor of leaving it to the superpowers.”

–IANS

rs/