U.S. President Joe Biden said that the risks of artificial intelligence have not yet been studied. However, in his view, companies must be responsible for the safety of AI products, writes Reuters.
In a meeting with science and technology advisers, Biden said that artificial intelligence can help in the fight against diseases and climate change. At the same time, it is important to address potential risks to society, national security and the economy, he added.
«In my view, technology companies must ensure that their products are safe before releasing them», the president said.
When asked whether AI is dangerous, Biden replied:
«That remains to be seen. Perhaps».
According to the president, social networks have already demonstrated the harm that powerful technologies can do without proper safety measures.
«In the absence of guarantees, we see impacts on mental health, self-image, feelings and hopelessness, especially among young people», Biden said.
He repeated the call for Congress to pass a bipartisan privacy bill. The initiative would impose limits on data collection online, ban targeted advertising to children, and make health and safety a priority in product development.
Before the meeting, AI company shares fell sharply. For example, shares of C3.ai Inc fell 24%. Shares of Thai security company Guardforce AI fell 29%, BigBear.ai down 16%, and SoundHound AI down 13%.
This is not the first time the administration has turned its attention to technology. In 2022, the White House released the AI rights bill, aimed at ‘protecting the American public’. The document advocates five principles:
- the safety and effectiveness of systems;
- data privacy;
- algorithmic safety from discrimination;
- notification and explanation of the purposes of using AI systems;
- human alternatives, consideration and opt-out from using algorithms.
Earlier in March, the technology ethics group urged the U.S. Federal Trade Commission to conduct an investigation into OpenAI.
In the same month, more than a thousand AI experts urged pausing the development of large language models for six months. Among the signatories were Elon Musk and Steve Wozniak.