AI companies will commit to safeguards in the US

The tech firms will need to share security information with other companies and the US government, and have to watermark AI-generated audio and visual content for easier identification.(Image: Representative Image)

Companies such as Microsoft, Google, and OpenAI among others have agreed they will commit to certain safeguards regarding Artificial Intelligence (AI), following a request by the White House.

According to Bloomberg, the companies have agreed to certain principles regarding the use and research of AI technologies but the agreement will expire once the US Congress passes AI legislation.

Also read | The US government announces voluntary 'Cyber Trust Mark'

The firms will agree to abide by eight principles according to a draft document. Under the agreement, the companies will allow independent audits from AI experts who will test their AI models.

They will also agree to invest in cybersecurity technologies, and allow third-parties to discover vulnerabilities. The companies will need to flag societal risks posed by AI, and abide to do research on how to reduce the societal risks of AI.

The tech firms will need to share security information with other companies and the US government, and watermark AI-generated audio and visual content for easier identification. The companies also agree to use AI systems for research in tackling societal problems.

As AI technology expands, the race to regulate it is on. The European Parliament has already passed a draft for the AI act.

India's Telecom Regulatory Authority (TRAI) has recommended an independent regulatory body to govern the use of AI. A digital data protection bill is already tabled in the parliament.

Also read | India, US commit to working towards making AI trustworthy, mitigating risks

Last month, the National Institute of Standards and Technology (NIST) launched a public working group to discuss the challenges and risks posed by AI.

US lawmakers have also proposed a "blue ribbon commission" to study the impact of AI.