OpenAI forms safety and security committee as it starts training new AI model

share on

twitter /
facebook /
linkedin /
- email
- telegram
- whatsapp
- wechat
- pinterest
- line
- snapchat
- reddit

The OpenAI board has formed a safety and security committee. The committee will be responsible for making recommendations to the full board on critical safety and security decisions for OpenAI projects and operations.

It will be led by directors Bret Taylor, Adam D'Angelo, Nicole Seligman and CEO Sam Altman.

The safety and security committee comes as OpenAI begins training for its next frontier model. For its first task, the team will evaluate and further develop OpenAI's processes and safeguards over the next 90 days. The committee will then present its findings to OpenAI's board.

Don't miss: OpenAI partners News Corp. for access to Wall Street Journal content

The board will then decide how to implement the recommendations in a manner that is consistent with safety and security.

OpenAI technical and policy experts Aleksander Madry (head of preparedness), Lilian Weng (head of safety systems), John Schulman (head of alignment Science), Matt Knight (head of security), and Jakub Pachocki (chief scientist) will also be on the committee.

In addition, OpenAI will retain and consult with other safety, security, and technical experts to support this work, including former cybersecurity officials, Rob Joyce, who advises OpenAI on security, and John Carlin.

Earlier in May, OpenAI announced a new AI model called GPT-40. The new model will be capable of realistic voice conversation and can interact across audio, vision and text in real time.

In a statement by OpenAI, it said that GPT-4o is able to accept any combination of text, audio and image as input and generate any combination of text, audio, and image outputs.

In addition, it can respond to audio inputs in as little as 232 milliseconds, with an average of 320 milliseconds, which is similar to human response time in a conversation, said OpenAI.

The new AI model has safety built-in, through techniques such as filtering training data and refining the model’s behaviour through post-training.

A new safety system was also created to provide guardrails on voice outputs.

Join us on 12 June 2024 for an exciting experience as Content360 makes its debut in Malaysia! Brace yourself to join the crème de la crème of the content marketing industry hailing from across the region. Immerse yourself in a dynamic atmosphere, and uncover the latest trends with thought leaders and solution providers from the realm of content.

share on