The White House confirmed US tech giants volunteering to conduct safety procedures before launching new AI tools.

The White House recently announced that it had received confirmations from nearly seven U.S. companies voluntarily committing to making sure artificial intelligence (AI) products released by each company stand by certain safety regulations. Part of the commitment requires some companies to have additional third-party oversight in order to make sure any form of commercial AI tech is run through an audit; however, details on who might audit these systems have not been provided.

Some of the companies that have agreed to meet President Joe Biden’s administration’s policies, requesting the further development of AI technology go through a set of safety protocols, include Microsoft, Meta, Google, and Amazon, among others.

In a statement by the White House, the four above mentioned tech giants, in addition to ChatGPT’s OpenAI and startups Anthropic and Inflection, have assured the government of carrying out security tests by independent experts to avoid huge risks like biosecurity and cybersecurity issues.

After a sudden surge in generative AI and the commercial investment in the industry with AI tools that could nearly show functions that were human-like, like writing and researching with near human capabilities, fascination and concern hit people simultaneously, with dangers like being able to trick people or spread disinformation starting to take hold.

Companies have therefore taken the step of consciously finding techniques that allow vulnerabilities to be reported to the systems in charge. The use of digital watermarking is one way companies help people distinguish between AI-generated texts or images and real ones.