Germany Takes AI Safety Threats Seriously

The German government is taking warnings about potentially uncontrollable artificial intelligence seriously, with the National Security Council having already examined the threat twice.

Deputy government spokesperson Steffen Meyer said Berlin did not want to focus solely on catastrophic scenarios, but also sought to harness AI’s potential in areas such as the economy, productivity and medicine. He also called for international solutions, including through the G7.

A Digital Affairs Ministry spokesperson described the threat posed by increasingly powerful AI as a “paradigm shift”. He warned of a scenario in which AI models could break out of testing environments and independently gain access to other systems.

Germany has recently established the National AI Safety Institute, which is tasked with identifying security-related risks associated with AI at an early stage.

The debate intensified after Anthropic CEO Dario Amodei renewed calls over the weekend to slow AI development. He was backed by OpenAI CEO Sam Altman and xAI chief Elon Musk.

(tasr, mja)