Labor is planning to require artificial intelligence companies to share the results of their road tests with authorities, replacing voluntary testing agreements with a statutory system. Peter Kyle, the shadow technology secretary, emphasized the need for greater transparency from tech companies, particularly in the wake of Brianna Gee’s murder.
Under Labor’s proposals, AI companies would be required to disclose their plans for developing AI systems and ensure safe testing under independent oversight. The testing agreement announced at the Global AI Safety Summit was supported by the EU and other countries, including the US, UK, Japan, France, and Germany.
During a visit to the United States, Kyle emphasized the importance of test results in providing independent scrutiny of cutting-edge AI technology. He stressed the need to ensure the safe development of technology that will have a significant impact on workplaces, societies, and cultures.
Tech companies that have agreed to test their models include Google, OpenAI, Amazon, Microsoft, and Meta. Kyle also highlighted the role of the British AI Safety Association in independently scrutinizing AI development.
“We are moving from voluntary regulations to statutory regulations,” Mr Kyle told BBC One’s Sunday with Laura Kuenssberg. We can find out what they’re testing for, so we know exactly what’s going on and where this technology is taking us.”
At the first Global AI Safety Summit in November, Rishi Sunak announced voluntary agreements with major AI companies such as Google and OpenAI. Under Labor’s proposals, AI companies would be required to disclose their plans for developing AI systems and ensure safe testing under independent oversight.
He added: “Some of this technology will have a profound impact on our workplaces, societies and cultures. And we need to ensure that its development occurs safely.”
Source: www.theguardian.com
