Meta, Google, and other advanced tech companies open the door to government oversight of AI
Elon Musk has warned that artificial intelligence may eventually replace all human jobs, so advanced companies will allow governments to test their AI tools for the first time.
Cutting-edge tech companies including Meta, Google DeepMind, and OpenAI have allowed regulators to test their latest artificial intelligence (AI) products before making them public, a move that officials believe will slow the race to develop systems that can compete with humans for jobs.
The testing agreement is supported by the EU and 10 countries, including the United States, Britain, Japan, France, and Germany. Among the leading AI companies that have agreed to test are Google, OpenAI, Amazon, Microsoft, and Meta.
British Prime Minister Rishi Sunak made the announcement on November 2 after a two-day summit in Bletchley Park, where an expert group, the US Vice President and a senior Chinese government official agreed that AI poses a serious danger to humanity, and Elon Musk warned that the technology could eventually replace all human jobs.
However, Sunak was forced to defend the voluntary nature of the testing agreement, as his government refused to introduce legislation to curb AI development.
Under the agreement, governments and AI companies agreed to work together to test the security of new AI models before and after their release. The large language models that underpin tools such as the ChatGPT chatbot will be tested in cooperation with governments for a number of dangers, including national security and harm to society.
The testing will be conducted under the guidance of new AI security institutes in the United States and the United Kingdom. Sunak noted that the British body positions itself as the “global center” of a multilateral initiative that does not involve China.
Sunak said that UN Secretary-General Antonio Guterres helped to gain international support for an expert panel to publish a report on the “state of AI science.”
“We should look at artificial intelligence more as a co-pilot than as something that will necessarily replace someone’s job,” the British Prime Minister added.
