On the AI Action Summit

We opened considering the AI Action Summit, https://www.elysee.fr/en/sommet-pour-l-action-sur-l-ia. Here are some notes and links, …

The British Press reporting has focused on the failure/refusal of the US/UK to sign the final declaration, https://www.theguardian.com/technology/2025/feb/11/us-uk-paris-ai-summit-artificial-intelligence-declaration

Others take a different slant,

1. https://www.business-standard.com/world-news/ai-action-summit-in-paris-calls-for-human-centric-approach-towards-ai-use-125021200513_1.html
2. https://www.reuters.com/technology/artificial-intelligence/europe-looks-embrace-ai-paris-summits-2nd-day-while-global-consensus-unclear-2025-02-11

As does the limited civic society impact,

1. https://about.make.org/articles-en/manifesto-call-for-a-worldwide-alliance-on-ai-and-democracy
2. https://www.lemonde.fr/idees/article/2025/02/04/l-appel-d-une-centaine-d-ong-l-ia-telle-qu-elle-est-developpee-alimente-un-systeme-d-exploitation-global_6531422_3232.html
3. https://www-lemonde-fr.translate.goog/idees/article/2025/02/04/l-appel-d-une-centaine-d-ong-l-ia-telle-qu-elle-est-developpee-alimente-un-systeme-d-exploitation-global_6531422_3232.html?_x_tr_sl=auto&_x_tr_tl=en&_x_tr_hl=en-US&_x_tr_pto=wapp

In particular the Commission, held up its position, https://www.politico.eu/article/virkkunen-stands-firm-against-american-pushback-against-eu-tech-laws

And https://www.reuters.com/technology/artificial-intelligence/quotes-eu-chief-von-der-leyens-ai-speech-paris-summit-2025-02-11

So what is the AI Act that the tech bros. object to?

1. https://artificialintelligenceact.eu
2. https://www.europarl.europa.eu/topics/en/article/20230601STO93804/eu-ai-act-first-regulation-on-artificial-intelligence
3. https://www.corporateeurope.org/en/2025/01/setting-rules-their-own-game-how-big-tech-shaping-ai-standards

In my conversation, we considered the issue of copyright law, and the ownership and sharding of the rules bases i.e. the data input to the training processes. We looked at the potential for conflict with the GDPR’s right of “freedom from profiling”. U.S. business and probably its government do not want strong regulation of AI.

https://www.reuters.com/technology/artificial-intelligence/europe-looks-embrace-ai-paris-summits-2nd-day-while-global-consensus-unclear-2025-02-11

The copyright disputes have their aspects of schadenfreude,

1. https://www.theguardian.com/technology/2025/feb/11/uk-copyright-law-consultation-fixed-favour-ai-firms-peer-says
2. https://futurism.com/openai-deepseek-permission-ai-stealing

Democratic Oversight of the models, reflects the profit/not for profit argument in the US and the need for the four freedoms and thus open source, noting that the Chinese alternative to OpenAI, Deepseek, is open source.

DeepSeek’s open-source revolution: A game-changer for AI development?

We agreed to use these insights in our open letter the special committee on democracy.

I wrote,

Apart from the issue of rules sharding and ownership, and the right to “freedom from profiling”, I’m interested in the way Google are using AI to summarise their findings. Where it occurs, I think it’s quite useful, particularly since they started to add citations, but does it undermine the principle of popularity. Google have also weakened this competitive advantage by allowing people to buy places in their search replies. What made and makes Google search so powerful is/was the basic in-list algorithm, even though it that caused difficulty when trying to distinguish between authority and popularity of the facts sought.

From this latter para, I made this https://www.linkedin.com/pulse/ai-open-source-digital-liberty-david-levy-ma0ve/ on Linkedin