Five things you need to know about the AI Safety Summit (and what comes next)

Hello,

Hello from Paris, where the AI Action Summit has just wrapped up. This global gathering brought together heads of state, tech CEOs, and civil society leaders to discuss the future of AI.

The past few days were filled with bold ideas, cautious optimism, and genuine frustrations about AI's direction — and questions about whether we're having the tough conversations needed to ensure that AI serves the common good.

The Summit was a mixed bag of progress and missed opportunities. Debates on the role of regulation took center stage, while AI's role in consolidating the power of the biggest tech companies lingered in the background, left largely unaddressed. And while there was plenty of talk, many pressing questions — like those around labor disruption and action beyond financial commitments — were often avoided or softened to prevent conflict.

Fortunately at Mozilla, we’re not afraid to have those difficult conversations. We've distilled the key takeaways from this year’s Summit, but talk alone isn’t enough.

Now is the time to push for AI that is open, accountable, and built for the common good. If you believe that AI should be trustworthy and serve all of us, sign Mozilla's petition now to join this movement.

1. The Summit spotlighted AI for the common good, not just profit.

Let’s start with the good. While AI often feels like a global race, where economic interests dominate and the public interest is left in the dust – this Summit struck a different tone. Public interest AI took center stage for a change.

Take the launch of Current AI, a new foundation launched at the Summit to advance AI in the public interest backed by the French government, or ROOST, a private initiative (backed in part by Mozilla!) focused on building open tools for online safety and content moderation.1 Government leaders like European Commission President Ursula von der Leyen and India’s Prime Minister Narendra Modi also highlighted the importance of openness in AI. The official Summit declaration included good ideas like championing transparency, resisting excessive market concentration, and promoting cultural and linguistic diversity.

Speaking of linguistic diversity: Mozilla’s Common Voice project was front and center, selected as one of 50 projects that lead the way in ethical AI.2

2. Unfortunately, it quickly turned into a game of ‘Who’s winning at AI?’

This is all great, but there’s a big BUT.

We really are in the middle of a global AI race that is unfolding in a tense geopolitical climate. The Summit made that as clear as ever. Keeping pace with the US and China became a big theme, with France and the EU announcing initiatives to mobilize €100+ billion investments in AI.3 There was loud pushback against regulation meant to ensure that AI works for the public good, with many arguing that any guardrails are slowing down innovation.4

And then there was the political friction. The US and the UK chose not to sign the official Summit declaration — deliberately refusing to put their name to a declaration that emphasized, above everything else, inclusivity and sustainability.5

3. This is just the beginning - the real work to make good on lofty promises starts now.

We should view this Summit as a checkpoint, not the finish line, on the long road to building a more open and trustworthy AI ecosystem. Whether it sparks lasting change or remains a one-off display of political ambitions depends on what happens next. Follow-through is what matters.

Now, it’s on government and industry leaders to turn words into action – to uphold their commitments to openness, transparency and trustworthiness, and invest in the initiatives launched at the Summit. It’s also on every one of us who cares about trustworthy AI to hold these leaders accountable, and to keep pushing for an AI ecosystem that’s not just innovative, but also fair, transparent, and built for the common good.

4. The Summit was more inclusive than previous iterations, but we need to raise the bar.

Compared to previous AI Summits – like the UK’s Bletchley Park convening in late 2023 – the French AI Action Summit took a big step forward for inclusivity, bringing in more voices from civil society communities, like Mozilla’s. This should be the benchmark for future AI Summits.

But there’s still more to do. Getting civil society in the room isn’t enough – non-profit organizations, activists, and community leaders need a real seat at the decision-making table. Including diverse perspectives and empowering those people to have their voices heard is not optional: it is a precondition if events like this actually want to catalyze meaningful change that benefits the public. If AI is going to work for everyone, then everyone needs to be part of shaping its future.

5. Together we can, and must, build AI that serves all of us.

Globally, it feels like we’re in that classic cartoon moment: We have run off the cliff, suspended in midair, legs still spinning, pretending there’s solid ground beneath us. And for now, we’re all playing along, pretending the ground is still there.

That’s exactly where AI governance stands today — racing forward without enough guardrails, while we hold our breath, hoping for stability. The risks are mounting: unchecked power and a widening gap between AI’s rapid development and our ability to ensure it serves the public good. We can either keep pretending everything is fine, or we can face reality and start building an AI future that puts people first — not just profit and power.

And yet, despite all this, I’m still optimistic. AI doesn’t have to be a dangerous cliff. It can be a bridge to open, accountable technology designed to benefit everyone. Maybe not tomorrow or next week, but with the right action, we can get there.

Thank you for everything you do for the internet and to build the digital future that we deserve.

Always yours,

Nabiha Syed
Executive Director
Mozilla

More Information:

1. Read more about ROOST on Mozilla’s blog and Current AI.

2. AI Action Summit / Paris Peace Forum: Find out more about the 50 Selected Projects.

3. Reuters: Macron signals investments of 109 billion euros in French AI by private sector. 9 February 2025.

4. Politico: JD Vance warns Europe to go easy on tech regulation in major AI speech. 11 February 2025.

5. Élysée: Statement on Inclusive and Sustainable Artificial Intelligence for People and the Planet. 11 February 2025. Guardian: US and UK refuse to sign Paris summit declaration on ‘inclusive’ AI. 11 February 2025.