The major players in the AI race – including Google, Microsoft, OpenAI, and Anthropic – have announced the formation of a new partnership, the Frontier Model Forum, a new industry body focused on ensuring safe and responsible development for frontier AI models.

The Frontier Model Forum will draw on the technical and operational expertise of its member companies to benefit the entire AI ecosystem, such as through advancing technical evaluations and benchmarks, and developing a public library of solutions to support industry best practices and standards.

So far, the companies have released the core objectives of the Forum, their criteria for membership, and what they aim to accomplish.

Core objectives include:

1. Advancing AI safety research to promote ‘responsible’ development of frontier models, minimise risks, and enable independent, standardised evaluations of capabilities and safety.

2. Identifying best practices for the responsible development and deployment of frontier models, helping the public understand the nature, capabilities, limitations and impact of the technology.

3. Collaborating with policymakers, academics, civil society and companies to share knowledge about trust and safety risks.

4. Supporting efforts to develop applications that can help meet society’s greatest challenges, such as climate change mitigation and adaptation, early cancer detection and prevention, and combating cyber threats.

Membership will be determined by the current forum, and is open to organisations that develop and deploy frontier models (as determined by the Forum), demonstrate a strong commitment to frontier model safety, and are will to participate in joint initiatives and supporting the development and functioning of the initiative.

According to a joint press release by the founding companies, the Forum will focus on three key areas to support the safe and responsible development of frontier AI models.

This includes identifying best practices and sharing this information across industry, government, civil society and academia, with a focus on safety standards and practices to mitigate potential risks.

Another area of focus in advancing AI safety research, first by identifying the most important open research questions on AI safety. The Forum aims to coordinate research to progress these efforts in areas such as adversarial robustness, mechanistic interpretability, scalable oversight, independent research access, emergent behaviours, and anomaly detection.

The Forum says there will be a strong focus on creating and sharing a public library of technical evaluations and benchmarks for frontier AI models.

Further, the Forum aims to facilitate information sharing among companies and governments by establishing secure mechanisms for sharing information regarding AI safety and risks, whilst following protocols for responsible disclosure.

Now just launched, the Forum is set to form an advisory board to guide its strategy and priorities. The current founding companies will establish a charter, governance and funding, and form an executive board, with plans to consult with governments and civil societies

The announcement marks a major collaboration among competing companies. Microsoft and OpenAI are directly competing against Google and Anthropic’s AI models. Further, Microsoft and Google are direct competitors in different ventures, most notably the Cloud industry.

Recommended

• What’s On Sale on the Dark Web?
• UK Universities’ Breached Credentials Found on Dark Web
• Police Crackdown on DDoS Services Users Following Europol Operation

The Forum comes not long after the EU announced it was in the final stages of approving its landmark AI Act, which is set to strictly regulate the development and deployment of AI models across different sectors and industries.

While AI companies have claimed to care about safety, and CEOs even at one point signed an open letter to pull the brakes on its rapid development and proliferation, there has been major pushback against regulation. CEOs signed another open letter addressed directly at EU legislation, saying that Europe risked being left behind if they regulated the emerging technology too closely. The letter urged for increased inclusion of industry experts to monitor the law, rather than lawmakers themselves.

Sam Altman, CEO of OpenAI, even reportedly lobbied against the EU AI Act to have his ChatGPT model not be classified as high-risk.

In the US, seven AI companies just agreed to safeguard spearheaded by President Joe Biden, which urged for security testing of AI models, watermarking AI-produced content, publicly reporting AI capabilities and limitations, and researching the associated risks of AI.

While the companies may be directly competing against each other, the announcement of the Forum could help companies better reconcile and comply with government regulation.