The White House announced on July 21, 2023, that major tech companies, including OpenAI, Amazon, Anthropic, Google, Inflection, Meta, and Microsoft, have committed to enhancing the safety of their artificial intelligence (AI) technology. They plan to achieve this by implementing features like watermarks on fabricated images. The watermarks aim to help users identify content generated by AI, preventing potential deception, fraud, and misinformation. This coming as the US Presidential elections for 2024 looms near and campaign season is about to begin.

The companies’ commitments are centered on three fundamental principles for AI’s future: safety, security, and trust. By developing robust technical mechanisms such as watermarking systems, they seek to create a reliable way for consumers to differentiate AI-generated content from authentic media.

Tech firms are also assessing potential biosecurity and societal risks by performing rigorous and comprehensive testing of current and developing AI systems. The overarching goal is to establish an easily accessible regime that enables people to discern AI-generated content in both audio and visual formats.

These measures are part of a broader effort to ensure responsible AI development and mitigate potential risks. However, implementation will require further technical work.

The White House has praised these commitments as a critical step towards developing trustworthy AI. However, there are concerns about whether tech companies will genuinely adhere to their voluntary pledges and support robust regulations in the future.

US President Joe Biden, in response to developing concerns, is working on issuing an executive order to ensure the safety and reliability of AI technology. This effort highlights the significance of AI governance and the need for responsible development in the fast-evolving technological landscape.

The discussions about watermarking AI-generated content also extend beyond the US, as evidenced by EU commissioner Thierry Breton’s meeting with OpenAI CEO Sam Altman in June 2023. This indicates that policymakers and industry leaders worldwide are exploring ways to address the challenges posed by AI-generated media.