AI's existential threat calls for immediate global action

Gladstone AI released a report emphasizing the critical need for immediate and comprehensive regulation of advanced artificial intelligence (AI) technologies following a commission by the U.S. State Department.

This report draws significant attention to Canada’s readiness to face the potential threats AI poses.

Titled "Defence in Depth: An Action Plan to Increase the Safety and Security of Advanced AI," the report reveals a dire scenario where unchecked AI systems could spiral out of control, posing an "extinction-level" threat to humans. In an interview with CBC's Power & Politics, Jeremie Harris, CEO of Gladstone AI explained the dangers.

"There's this risk that these systems start to get essentially dangerously creative... They're able to invent dangerously creative strategies that achieve their programmed objectives while having very harmful side effects," he said.

The report also highlights the risks of AI systems being weaponized. Harris shared concerns about AI-powered cyber threats with CBC, saying, "Imagine the same system... you're asking it to carry out a massive distributed denial of service attack or some other cyber attack."

These revelations have put the spotlight back on Canada's efforts in AI safety and regulation. The current Artificial Intelligence and Data Act (AIDA), part of Bill C-27, is under scrutiny.

Conservative MP Michelle Rempel Garner criticized the bill, telling CBC News, "C-27 is so out of date." She pointed out the need for more current regulations in light of new AI advancements like OpenAI's ChatGPT.

The U.S. State Department, while confirming the report's commission, clarified that it doesn't represent the views of the U.S. government.

The report’s urgency echoes through global discussions on AI regulation. It proposes critical actions, including export controls, new laws and comprehensive AI development frameworks.

Harris emphasized the importance of these steps to U.S. officials, explaining: "By the time AIDA comes into force... Frontier AI systems will have been scaled hundreds to thousands of times beyond what we see today."

Source: Newsroom