Back to All Events

Generative AI September Learning Lab: AI Policy

Tech & AI are in a profound state of flux and a default response is to call for policy changes. But what does that actually mean? What should we hope policy can accomplish? And what role can funders play in supporting good policy?

This session focused on what political tools are available to us, what we have learned from policy making efforts thus far, and how to prevent AI policy from being a game of the powerful rather than a meaningful step to protecting peoples’ rights.

Our speakers:

  • Fanny Hidvégi, Policy Director at the AI Collaborative and former director of Access Now Europe, shared analysis on approaches to AI policy that centers human rights and what we have learned from the EU AI Act about pursuing and securing meaningful policy, despite industry influence and an increasing securitisation agenda.

  • Ami Fields-Meyer, Senior Fellow at the Harvard Kennedy School and former Senior Policy Advisor to Vice President Kamala Harris at the White House, shared insights on adopting a more assertive and nimble political strategy to close the power gap between the interests of tech companies and the needs of ordinary people.


To keep the discussion as candid and free-flowing as possible, we opted not to record the session this time. If you missed it, keep reading to get an overview of the key points.


When frameworks and whitepapers aren’t enough

In this session, both speakers emphasised an urgent need for change in how we construct AI policy. Here’s an overview of what was covered:

Civil society is in an uphill struggle to shape public perception of AI. Tech companies have consolidated a great deal of power, and therefore have undeniably deep pockets for influencing public discourse on AI. Civil society is comparatively under-resourced; their voices fail to cut through all the noise because they don’t have the same influencing power, and they struggle to co-ordinate around messaging — partly because there are so many issue areas that require focus, and partly because the strategy so far has been to prioritise policy creation over driving against the dominant narrative that AI is an inevitability and inherent good.

There is a huge gulf between the priorities of technologists developing AI, and everyday citizens. Right now, there are no real political consequences to ignoring the concerns of regular people, meaning politicians will vote against civil protections in lieu of bolstering perceived progress in tech innovation. The policy space is therefore flooded with industry voices and needs to be diversified.

Current policies are risk-based and prohibitive rather than being geared towards public interest and human flourishing. Fanny noted the lack of resources that regulators have to enforce their own policies, which with AI have so far been lacking in substance, focussing on how to restrict company activity over how to promote human rights. Many tech policies also rely on a risk-based framework, which requires entities to self-assess on the risks that they create, which isn’t the best catalyst for accountability.

What can be done:

Civil society needs to be empowered to engage more meaningfully in the politics of AI. This means support from philanthropy to help level the playing field:

  • Funding can allow civil society to build influencing power and gain more influence over the narratives that dominate the space.

  • The production of whitepapers and frameworks isn’t enough; there is a need for more formalised sharing of research and resources, in order to actually co-ordinate on messaging, and counter the narratives spun by the tech industry — stories are often more powerful than raw research.

Right now, funders can do a lot to facilitate a mindset shift on how this kind of policy work should be approached:

  • There is already a lot of valuable research being produced about the harms and risks of generative AI — we need a mechanism to disseminate this to public interest groups such as labour unions, so they have the knowledge they need to put pressure on politicians to make change.

  • Fanny mentioned how the French Summit have already made an effort to re-situate policy discussions around human rights, rather than an exclusive focus on prohibitions and restrictions. This shift needs to be encouraged, to make room for policy that promotes human flourish and corporate accountability.