Today sees the launch of the Robust Open Online Safety Tools (ROOST) initiative at the AI Action Summit in Paris.
ROOST is a new non-profit organisation incubated at the Institute of Global Politics at Columbia University 'that brings together the expertise, resources, and investments of major technology companies and philanthropies with the aim of building scalable, interoperable safety infrastructure suited for the AI era.
The summit’s priorities include ‘ensuring that AI is open, inclusive, transparent, ethical, safe, secure and trustworthy.’
According to Amanda Brock, CEO of OpenUK, the launch is part of a broader commitment to open source technology, that is critical to Europe:
“From the get-go, as we would expect in Macron’s France, unlike past summits, the AI Action Summit has referenced open source and the open source community.
Its outputs will be grounded in three main principles of science, solutions - focusing on open AI models- and policy standards.”
She notes that advocating for governance through ‘open source digital tools that manage AI’ as opposed to regulation is something OpenUK has been doing for some time."
“It’s the only plausible route to governance and safety whilst enabling innovation. The UK’s AI Safety Institute led the way — open sourcing its ‘Inspect’ LLM Evaluation Platform back in May — for today’s launch of ROOST’ at the Summit."
According to Brock, this approach is key to the future management. governance and safety of AI.
“It sees the European Commission increasingly out of step, as it implements its cumbersome and complex AI Act in juxtapositions to the rest of the world.”
ROOST will offer free, open-source, and easy-to-use tools to detect, review, and report child sexual abuse material (CSAM); leverage large language models (LLMs) to power safety infrastructure; and make core safety technologies more accessible and more user friendly.
With dedicated technical teams providing hands-on support, ROOST will meet organisations where they are, helping them integrate robust safety measures while continuing to innovate.
Dr Laura Gilbert, OpenUK Advisory Board, asserts that when we make AI tools and safety frameworks open source, we create a global laboratory for innovation that simultaneously builds trust through transparency.
“This approach may allow us to maintain the rapid pace of AI advancement whilst helping to ensure appropriate safeguards are in place and well understood.
The technology's immense potential can only be properly realized when development and safety protocols are accessible to all, enabling collaborative governance that transcends national boundaries.”
ROOST's founding partners include:
- Eric Schmidt,
- Discord,
- OpenAI,
- Google,
- Roblox,
- John S. and James L. Knight Foundation,
- AI Collaborative,
- Patrick J. McGovern Foundation, and
- Project Liberty Institute.
ROOST also benefits from the support and expertise of a wide set of partners from the fields of AI, philanthropy, academia, open source, child safety, and countering violent extremism.
To date, ROOST has raised more than $27 million for its first four years of operations from a range of leading philanthropies and top technology companies. ROOST aims to accelerate its growth and expand its offerings to as many organisations as possible.
According to Brock, ensuring our AI futures don’t sit in a handful of companies is an underlying theme across all conversations in this Europe-based summit.
“The real question will be whether the US and China can be persuaded to sign up to the end-of-summit statement.
For the UK it's clear we must build domestic capabilities whilst collaborating with international companies to build the infrastructure critical to the UK's success. To be world-class, we need infrastructure in the UK stewarded with standards that create global norms and governance through open source tooling."
Further announcements at the Summit will include the launch of the ‘Current AI’ Foundation, funded by ‘willing nations’ to the tune of €400m on launch, and include data commons and ethical AI practices as necessary enablers of the democratisation of AI.
Would you like to write the first comment?
Login to post comments