Center for AI and Digital Policy Europe’s Post

It was a great pleasure to participate this morning in a panel on AI regulation in Europe organised in the framework of the QUO VADIS Conference in Athens, together with MEP Brando Benefei Peter Kimpian Alessandro Mantelero Vagelis Papakonstantinou One key take away is that we need to think in sync the application of the EU AI Act, DSA and GDPR, together with that of Council of Europe Convention on AI, human rights, democracy and the rule of law and the Council of Europe Convention 108+ on data protection. 1. because this is the only way we can ensure a rights-based approach to AI regulation which is the trademark of the "European way" 2. because EU Member States will have to apply both European layers of rules, on top of the national ones, a great baklava -- thank you Peter Kimpian for the metaphor ;), so they need to get ready now 3. because we can expect in practice the integration of the data protection impact assessment within the matrix of eg Art. 9 EU AI Act risk assessment system or Art. 27 fundamental rights impact assessment. On a side note, as I was asked about the regulation of AI systems used for national security purpose, I would say: - do not look at the cake only from the perspective of the AI Act or the CoE Convention (spoiler: we can expect a battle of interpretation on what's in & what's out) - also look for example at the case law of the European Court of Human Rights eg its recent judgment (28 May 2024) in the case of Pietrzak, Bychawska-Siniarska and Others v. Poland (applications nos. 72038/17 and 25237/18) https://lnkd.in/eemZ_WSZ The case concerned a complaint by five Polish nationals (kuddos to Barbara Grabowska-Moroz, Panoptykon Katarzyna Szymielewicz and Mikołaj Pietrzak) about a Polish legislation authorising a secret-surveillance regime covering both operational control and the retention of communications data for possible future use by the relevant national authorities (possibly feeding AI systems used for national security purpose). The Court held that: ❎ the mere existence of the relevant legislation (secret nature and wide scope of the measures provided for in the legislation + lack of effective review) constitutes in itself an interference with the complainants' Article 8 ECHR rights. 👉 lightened burden of proof: applicants do not have to prove that surveillance actually took place. Status of victim automatically granted. ❎ insufficient safeguards against excessive recourse to surveillance and undue interference with individuals’ private life ❎ no proper judicial review mechanism ❎ retention of communications data in a general and indiscriminate manner for possible future use by the relevant national authorities, without any review mechanism by independent body = not limited to what is “necessary in a democratic society”

Today, we joined European Digital Rights and nine civil society organisations in calling the European Commission to scrutinise and reconsider its re-confirmation of data transfer adequacy with Israel. This decision raises red flags, especially in light of the International Court of Justice (ICJ)’s ruling on a plausible genocide in Gaza. We call for more transparency on how Israel’s privacy and data protection standards align with those of the EU.

It's time for the #EuropeanCommission to reassess its data policy transfer with Israel. Recent revelations linking data biometric practices to AI targeting have raised serious concerns about compliance with #EUlaw. Specifically, data related to #Lavender indicates potential misuse that conflicts with established data protection standards. In light of Israel's ongoing actions in Gaza and its digital occupation of Palestine, it's crucial for the Commission to take a stand. The indiscriminate personal data transfers and building of mass surveillance capabilities must not be overlooked. These practices, including the creation of biometric databases and AI-generated kill lists, raise significant ethical and legal implications. As noted by MARWA FATAFTA, MENA POLICY AND ADVOCACY DIRECTOR AT ACCESS NOW, Israel cannot be seen as a data protection haven.

STRASBOURG, March 13 (Reuters) - Europe moved closer to adopting the world's first #artificial #intelligence #rules on Wednesday as EU lawmakers endorsed a provisional agreement for a technology whose use is rapidly growing across a wide swathe of industries and in everyday life. The legislation will #regulate high-impact, general-purpose AI models and high-risk #AI #systems which will have to comply with #specific #transparency #obligations and EU copyright laws. It #restricts governments' #use of real-time #biometric #surveillance in public spaces to cases of certain crimes, prevention of genuine threats, such as terrorist attacks, and searches for people suspected of the most serious crimes.

🌍 Today, with 11 CSOs, we have written to the EU Commission urging further scrutiny over its decision re-affirming Israel's data protection adequacy status - enabling unrestricted flow of data between Israel and the EU. Despite the presence of many red flags, in January of this year, the EC, after a review, concluded that Israel’s privacy and data protection standards align with those of the EU, reaching the "essentially equivalent" standard. We seek clarifications and explanations. In our letter we outline six critical legal considerations questioning the review of the European Commission of Israel's: current status of rule of law, privacy and data protection framework, provisions on national security and mass surveillance practices, onward transfers beyond its internationally recognised borders. Finally, we question the EC's review of "Israel's respect of international human rights norms and standards" (especially in light of the current accusations of genocide in Gaza).

Synthetic media is any type of media content (text, image, video, sound) that has been fully or partially generated using artificial intelligence. While non-synthetic media is created in the physical environment by humans, synthetic media is generated in the digital world by AI algorithms. It is significantly influential in the current landscape, posing challenges for law enforcement and demanding new standards in media literacy and critical thinking. Its rapid evolution and exploitation for malicious activities highlight the need for dynamic research, training, and international collaboration to tackle its threats effectively.

With the temperatures soaring near 49C/120F in the UAE, it’s a good time to find a cool spot and bury into a good read. I’m currently reading the INTERPOL Report “BEYOND ILLUSIONS – UNMASKING THE THREAT OF SYNTHETIC MEDIA FOR LAW ENFORCEMENT”. This report confirms what I’ve experienced, that is, with the rapid advancement of AI technology, synthetic media is becoming an influential mode of content. Its ease of availability has enabled its exploitation by criminals for malicious activities. Currently, synthetic media files possess the ability to deceive human perception, making it challenging to determine their authenticity. As AI continues to evolve, these artificially generated media will become increasingly sophisticated, posing significant challenges for law enforcement agencies in identifying, locating, and conducting forensic analyses of such content once criminal elements are arrested. This report on synthetic media and its implications for law enforcement offers a thorough introductory understanding of this pattern and trend. Based on the research and analysis of the current synthetic media landscape, some of the takeaways for me include: • Synthetic media isn’t new.  It’s growing at exponential rates and tackling the threats posed by it requires dynamic research and training in this landscape to understand it better and find detection solutions.  • If you are a law enforcement agency, you need to start investing funds, personnel, and training to this most serious threat if you haven't already. • To combat this threat effectively, law enforcement agencies must gain a comprehensive understanding of the multifaceted nature of synthetic media, including synthetic audio, text, and identifications; how they are created and distributed; and the huge impact if left unchecked. • Law enforcement requires a holistic solution to investigate digital media files efficiently and effectively in this synthetic content era. • The rise and increasing ease of AI-generated content calls for a collaborative approach between global law enforcement, the private sector, and academia to stay at the forefront of synthetic media forensics and investigations through ongoing vigilance, continuous technological advancements, and international collaboration. INTERPOL does a fantastic job with this report explaining and helping the reader understand the synthetic media landscape and gain some insight to help tackle the threats posed by it. Thank you Mark Nuttall for your brilliant contribution! 👍

Argentina’s AI-generated “Minority Report” Argentina’s security forces have announced plans to use artificial intelligence to “predict future crimes” in a move experts have warned could threaten citizens’ rights. The country’s far-right president Javier Milei this week created the Artificial Intelligence Applied to Security Unit, which the legislation says will use “machine-learning algorithms to analyse historical crime data to predict future crimes”. It is also expected to deploy facial recognition software to identify “wanted persons”, patrol social media, and analyze real-time security camera footage to detect suspicious activities. — The Guardian #argentina #javiermilei #AI #ML #machinelearning #predictiveanalysis #data #datacollection #surveillance #security #crime #policestate

Important Q&A by @hrw about some of the surveillance technologies, AI, & other digital tools used by Israel to decide who or what to attack in Gaza There are 4 key tools - all premised on systematic & invasive surveillance of Palestinians: 1) Use of mobile phone tracking to monitor the movements of Palestinians from parts of northern Gaza. Reports put the number of phones tracked at over 1 million (basically anyone with cell). But is it about ensuring safe evacuations or rather overseeing permanent transfer? 2) "Lavender" uses machine learning to assign residents of Gaza a numerical score relating to the suspected likelihood that a person is a member of an armed group. Basically algorithm decides if you look/behave like a Hamas member & assigns you a score - & this score probably can lead to your death. An Orwellian Big Data Russian roulette 3) The "Gospel" uses an algorithm to process surveillance data to generate lists of targets. It seemingly identifies nonhuman targets, including what it considers military targets; family homes of suspected militants; and “power targets” (basically civilian structures attacked to create a shock) 4) “Where’s Daddy” (IDF's frat boy morbid humor) is a tool that uses mobile phone location tracking to notify Israeli military when people who have been marked as military targets enter a specific location, such as a family home. I wonder if this explains why so many families are being decimated. Behind all this fancy tech (which is likely coming to a country near you as Israel exports it) is a basic fact: it is the digitalization and mass data enablement of genocide under the guise of "big data targeting". There is nothing sophisticated about this system as Gaza shows. We need to better understand these algorithms & technologies. One of the reasons why @ArabReform_ARI is partnering with Logic(s) to offer 4 fellowships to Palestinian journalists to critically examine the impact of technology on Palestinian lives https://lnkd.in/eTWN4nRm https://lnkd.in/e5Tv9cqA