A government-commissioned paper warns that there is a “clear and urgent need” for the US government to take action because rapidly advancing AI might lead to weaponization or loss of control, which could endanger humankind. “The rise of advanced AI and AGI has the potential to destabilize global security in ways reminiscent of the introduction of nuclear weapons,” says the report that TIME Magazine got, named “An Action Plan to Increase the Safety and Security of Advanced AI.”
In a report published by Gladstone AI Inc., it was stated that the US government must intervene immediately due to the increasing danger that AI is posing to national security due to its potential for weaponization and loss of control. The report went on to say that the continual proliferation of AI capabilities only makes these dangers worse.
Over thirteen months, researchers consulted about two hundred individuals, including representatives from the governments of the United States and Canada, prominent cloud providers, groups concerned with the safety of AI, and computer and security specialists, to formulate the report’s proposed intervention blueprint.
There is concern that “the rise of advanced AI and AGI has the potential to destabilize global security in ways reminiscent of the introduction of nuclear weapons,” according to the paper “An Action Plan to Increase the Safety and Security of Advanced AI.”
First, the goal is to put in place temporary protections for advanced AI, and then they will be formalized into law. The next step would be to make the safeguards global.
According to TIME, some potential steps include establishing a new AI agency to limit AI’s processing power, mandating that AI companies seek government approval before releasing new models that surpass a specific threshold, and possibly banning the disclosure of AI models’ inner workings, similar to open-source licensing. The research also suggested that the government should make it harder to make and export AI chips.
