Copy
The EuropeanAI Newsletter
Covering Artificial Intelligence in Europe

Welcome to the EuropeanAI Newsletter covering the European AI and technology ecosystem. If you want to catch up on the archives, they're available here.

You can share the subscription link here

The Council has agreed its position on the Digital Markets Act (summarised here) and on the Digital Services Act (summarised here).

A bill proposed by NY City Council would require that “automated employment decision tools” (e.g. algorithms) are audited for bias prior to use. It would also require users to be notified if these tools are being used. While many steps removed from the AI Act (i.e. a horizontal, risk-based approach to regulating AI), it is a step toward thinking about how to govern algorithms whose impact on individuals and society has the potential to be both negative and difficult to reverse.

The UK’s Cabinet Office’s Central Digital and Data Office and the Centre for Data Ethics and Innovation propose a standard for algorithmic transparency for public bodies and government departments.

In a refreshing change of pace,
some CEO’s have embraced the AI Act despite its potential financial burden (see more discussion on the hypothetical financial burden, and why it is sometimes overstated, here).

Policy, Strategy and Regulation

A narrowing of AI? The changing definition of AI in the AI Act
by Daniel Leufer, Access Now

According to a leaked “progress report” (privy to Access Now) on the AI Act presented to the EU Council, the Slovenian Presidency plans to present a first draft compromise at the end of November. The progress report gives a high-level overview of proposed amendments, including a proposal to narrow the definition of “AI systems” in Article (3)(1), and the list of “Artificial Intelligence Techniques and Approaches” in Annex I.

In response, digital rights groups Access Now and AlgorithmWatch
published a joint letter last week warning against any attempt to narrow the definition’s scope. While other stakeholders have criticised the broadness of the AI Act’s definition of AI (such as the consultation response from the Confederation of Artificial Intelligence Laboratories in Europe), Access Now and Algorithm Watch warn that any narrowing would undermine it. The joint letter outlines that narrowing the definition of AI to only include more advanced techniques like machine learning would exclude a range of systems that use more rudimentary techniques, but pose an equally high risk to fundamental rights. Moreover, such a narrowing would actually disincentive the use of cutting-edge techniques, as providers could avoid obligations imposed under the AI Act by using simpler techniques that are not in scope.


Notes in the Margin: Today, 114 civil society organisations published a joint statement to ask for changes to the AI Act to ensure that it is an effective instrument for protecting fundamental rights. The statement was drafted by European Digital Rights (EDRi), Access Now, Panoptykon Foundation, epicenter.works, AlgorithmWatch, European Disability Forum (EDF), Bits of Freedom, Fair Trials, Platform for International Cooperation on Undocumented Migrants (PICUM), and ANEC.


Proposal for a Regulation on Political Advertisement 

The European Commission has proposed new legislation to regulate political advertisements published in or targeted towards individuals in EU member states. The proposed regulation presents transparency measures, alongside data protection, retention, and publication requirements, for entities involved in the production and sharing of political advertisements. Political advertisers are defined as those involved in disseminating a message on behalf of a political actor that is liable to influence voting behaviour or electoral outcomes. 

The new legislation requires the use of  transparency notices (that identify the political nature of the ad, its sponsors, and its wider context to the general public) to be included in or retrievable from political advertisements. However, it is not clear what the latter would mean in practice. The proposed regulation increases provisions to protect personal data used in targeting or amplification of political advertisements. For example, Art 12 (1-2) proposes a prohibition on certain situations where targeting and amplification techniques involve the processing of personal data in the context of political advertisement. Some additional requirements are needed for the targeting and amplification of political advertisements where personal data is processed beyond the transparency notice in Art 7, such as, “a reference to effective means to support individuals exercise their rights under Regulation (EU) 2016/679 (GDPR)”.
 

Ecosystem

ADRA is ready to support the EU AI ecosystem

Earlier this summer, the European Commission and the European Partnership on AI, Data and Robotics (ADRA) signed a memorandum of understanding where the parties agree that ADRA will leverage EUR 1.3 bn of public investment through the Horizon Europe programme and match this with EUR 1.3bn of private investment until 2030. In pursuit of this ambition, ADRA has officially started its work with a partnership launch event last week. ADRA is the main Partnership for expertise and research innovation in value-driven and trustworthy AI, Data, and Robotics within the EU. It exists to bridge silos between key stakeholders, mobilise the ecosystem, and to expand deployment and adoption of these technologies throughout the EU. In doing so, it contributes to the EU’s wider strategic objective of leading the global development of robust and trustworthy AI, data, and robotics.
 

Enjoy learning about Europe? Share the subscription link with friends, colleagues, enemies...

Contact Charlotte Stix at:
www.charlottestix.com
@charlotte_stix


Ben Gilburt  co-wrote this edition.
Interesting events, numbers or policy developments that should be included? Send an email!

Disclaimer: this newsletter is personal opinion only and does not represent the opinion of any organisation.

Copyright © Charlotte Stix, All rights reserved.
Twitter Twitter
Website Website
Email Email






This email was sent to <<Email Address>>
why did I get this?    unsubscribe from this list    update subscription preferences
EuropeanAI · 16 Mill Lane · Cambridge, UK CB2 1SB · United Kingdom