OpenAI Implements Safety Measures, Board Can Reverse AI Decisions

December 19, 2023: OpenAI has introduced a comprehensive safety framework for its cutting-edge AI models. This move is significant as it empowers the company’s board to override decisions made by executives on safety matters.

This development, announced on the OpenAI website, reflects the company’s commitment to deploying technology responsibly, especially in sensitive areas like cybersecurity and nuclear threat management.

Backed by tech giant Microsoft, OpenAI has stated that it will only release its latest innovations if they are assessed as safe in critical domains. The firm is also forming an advisory group tasked with evaluating safety reports, which will then be forwarded to OpenAI’s executives and board members for review.

While the executives are responsible for initial decisions, the board holds the authority to reverse these decisions if necessary.

This initiative comes at a time when the AI community and the public are increasingly aware of the potential risks associated with advanced AI technologies.

Since the launch of ChatGPT a year ago, there have been growing concerns about AI’s ability to disseminate false information and manipulate human behavior.

The technology’s capabilities, ranging from composing poetry to crafting essays, have been both admired and scrutinized.

Earlier this year, AI experts and industry leaders signed an open letter urging a six-month halt in the development of AI systems more advanced than OpenAI’s GPT-4.

This highlights the apprehensions surrounding AI’s impact on society.

Supporting this sentiment, a Reuters/Ipsos poll in May revealed that over two-thirds of Americans are worried about AI’s adverse effects, with 61% believing it could pose a threat to civilization.