Luiza Jarovsky’s Post

🚨 [AI & PRIVACY] Singapore's Data Protection Authority publishes its "Proposed Guide on Synthetic Data Generation," and it's an interesting read for everyone in AI & privacy. Info & quotes: ➡ Among other topics, the proposed guide discusses a five-step approach to generating synthetic data, especially in the context of a Privacy-Enhancing Technologies (PETs) perspective: ➡ Step 1: Know your data "Before embarking on any synthetic data project, it is necessary to have a clear understanding of the purpose and use cases of the synthetic data and the source data that the synthetic data is to mimic. This will help to determine whether use of synthetic data might be relevant and identify the possible risks of using the synthetic data." ➡ Step 2: Prepare your data "To ensure that the synthetic data can meet the business objectives, organisations need to understand and identify the trends, key statistical properties, and attributerelationships in the source data that need to be preserved for analysis e.g., identify relationships between demographic characteristics of population and their health conditions." ➡ Step 3: Generate synthetic data "There are many different methods to generate synthetic data, for example, sequential tree-based synthesisers, copulas, and deep generative models (DGMs). Organisations need to consider which methods are most appropriate, based on their use cases, data objectives, and types of data. (...)Thereafter, organisations may consider splitting the source data into two separate sets e.g., 80% as training dataset, and 20% as control dataset for assessing re-identification risks of the synthetic data." ➡ Step 4: Assess re-identification risks "After the synthetic data is generated and utility measurement is assessed to be acceptable, organisations should assess and perform the re-identification risk assessment based on their internal acceptance criteria. (...). As synthetic data generally does not replicate its training data points, re-identification risk cannot be deduced directly from scrutinising whether the generated synthetic data contains any personal data." ➡ Step 5: Manage residual risks "In this final step, organisations should identify all potential residual risks and implement appropriate mitigation controls (technical, governance, and contractual) to minimise the identified risks. These risks and controls should be documented and approved by the management and key stakeholders as part of the organisation’s enterprise risk framework." ➡ It's an interesting guide, especially for those involved with AI development and Privacy-Enhancing Technologies (PETs). There is much more to read in this guide, make sure to check it out below. ➡ To stay up to date with the latest developments in AI policy & regulation, join 29,300+ people who subscribe to my weekly newsletter (link below). #AI #Privacy #PETs #PrivacyEnhancingTechnologies #AIpolicy

🚨 [AI REGULATION] A new AI law was introduced in the US to combat deepfakes, increase AI transparency, and empower creators. Here's what you need to know: ➡ The Content Origin Protection and Integrity from Edited and Deepfaked Media Act (COPIED ACT) was introduced last week by Senators Maria Cantwell (D-Wash.), Chair of the Senate Commerce Committee, Marsha Blackburn (R-Tenn.), member of the Commerce Committee, and Martin Heinrich (D-N.M.). ➡ According to the official release, among its highlights are: ➵ Creating transparency standards ➵ Putting journalists, artists, and musicians in control of their content ➵ Giving individuals a right to sue violators ➵ Prohibiting tampering with or disabling AI provenance information ➡ In this context, the COPIED ACT, among other provisions: "➵ Requires developers and deployers of AI systems and applications used to generate covered content (digital representations of copyrighted works) to give users the option to attach content provenance information within 2 years; ➵ Prohibits removing, altering, tampering with, or disabling content provenance information, with a limited exception for security research purposes; ➵ Prohibits the use of “covered content” (digital representations of copyrighted works) with content provenance to either train an AI- /algorithm-based system or create synthetic content without the express, informed consent and adherence to the terms of use of such content, including compensation;" ➡ It was endorsed by various alliances and groups of creative professionals, such as SAG-AFTRA, Nashville Songwriters Association International, The Recording Academy, National Music Publishers' Association, Recording Industry Association of America® (RIAA), National Newspaper Association, National Association of Broadcasters, Artist Rights Alliance, and more. ➡ These are interesting developments in the context of AI regulation, although we still need to wait and see if this bill will survive the legislative process. Find all relevant links below. 🚨 Additional resources on AI regulation & compliance ➵ To learn more about AI policy & regulation, join 29,000 people who subscribe to my weekly newsletter (link below). ➵ To upskill and obtain AI training, check out our AI Bootcamps at the AI, Tech & Privacy Academy (link below). #AILaw #AIRegulation #AIGovernance #AICompliance #Deepfakes

🚨 [DARK PATTERNS IN 2024] If you are interested in dark patterns & deceptive design, you CAN'T MISS these four 2024 reports. Read, bookmark & share: 1. The International Consumer Protection and Enforcement Network’s (ICPEN) annual review 2024. ➡ Quote: "Of the 642 traders swept in the ICPEN sweep, 75,70% made use of at least one dark pattern, and 66,82% made use of multiple dark patterns when marketing their subscriptions (see Figure 1). As such, out of all the traders subject to this sweep, only 156 traders were not found to employ any dark patterns to sell their subscription offerings." ➡ Read: https://lnkd.in/e9Y2Z3JR 2. The 2024 Global Privacy Enforcement Network (GPEN) Sweep 2024. ➡ Quote: "On average, the websites and apps examined used interface interference (e.g., language and visual tools) to influence users to select less privacy-protective options in 43% of interactions. 7 41% of websites and apps asked sweepers to make privacy choices when they first engaged with the websites and apps. 70% of those (or 31% of all websites and apps swept) made the less privacy-protective options easier to select." ➡ Read: https://lnkd.in/eb2_9Dva 3. Office of the Privacy Commissioner of Canada/Commissariat à la protection de la vie privée du Canada Sweep Report 2024 - Deceptive Design Patterns. ➡ Quote: "Specifically, OPC sweepers found that 33% of privacy policies were very difficult to read (compared to a global average of 20%). In addition, the privacy policies reviewed were found to be very long (76% were over 3,000 words, compared to a global result of 55%)" ➡ Read: Link: https://lnkd.in/ePt4z4XN 4. noyb.eu's Consent Banner Report ➡ Quote: "The purpose of this report is to offer a comprehensive account of the EDPB taskforce’s report findings for each violation, compared with the positions taken by national DPAs in guidance documents." ➡ Read: https://lnkd.in/eq2K3YnS 5. My academic article "Dark Patterns in Personal Data Collection: Definition, Taxonomy, and Lawfulness." ➡ Reports 1, 2, and 3 above were based on the OECD taxonomy, which cited my academic paper on dark patterns in privacy. ➡ Read: https://lnkd.in/dmY623tP 6. My newsletter ➡ In my weekly newsletter (29,000+ subscribers), I've covered dark patterns in privacy multiple times (see the archive). ➡ Read my article with examples here: https://lnkd.in/eUFUDuY6 - ♻️ Share this post and help your network stay up to date with the latest dark patterns reports. #darkpatterns #deceptivedesign #privacy #dataprotection

🌱 [AI ACT & ENVIRONMENTAL PROTECTION] Does the EU AI Act consider AI's environmental impact? Does it support environmental sustainability? Here's what you need to know: 1. [SUBJECT MATTER] Article 1, dealing with "Subject Matter," highlights environmental protection among the core values being protected: "The purpose of this Regulation is to improve the functioning of the internal market and promote the uptake of human-centric and trustworthy artificial intelligence (AI) while ensuring a high level of protection of health, safety, fundamental rights enshrined in the Charter, including democracy, the rule of law and environmental protection, against the harmful effects of AI systems in the Union and supporting innovation." 2. [STANDARDS] Article 40.2 dealing with "Harmonised Standards and Standardisation Deliverables" establishes that energy consumption must be taken into consideration: "(...) The standardisation request shall also ask for deliverables on reporting and documentation processes to improve AI systems’ resource performance, such as reducing the high-risk AI system’s consumption of energy and of other resources during its lifecycle, and on the energy-efficient development of general-purpose AI models. (...) 3. [CODES OF CONDUCT] Article 95.2 establishes that the AI Office will facilitate codes of conduct focusing on environmental sustainability: "The AI Office and the Member States shall facilitate the drawing up of codes of conduct concerning the voluntary application (...) (b) assessing and minimising the impact of AI systems on environmental sustainability, including as regards energy-efficient programming and techniques for the efficient design, training and use of AI;" 4. [ENVIRONMENTAL REVIEW] Article 112.2, dealing with "Evaluation and Review," establishes that environmental sustainability must be taken into consideration when reviewing the codes of conduct: "By 2028 and every three years thereafter, the Commission shall evaluate the impact and effectiveness of voluntary codes of conduct to foster the application of the requirements set out in Chapter III, Section 2 for AI systems other than high-risk AI systems and possibly other additional requirements for AI systems other than high-risk AI systems, including as regards environmental sustainability." 5. [GENERAL PURPOSE AI MODELS] Annex XI establishes that all providers of general-purpose AI models must share the "known or estimated energy consumption of the model" ♻️ Share this post and help your network learn about the EU AI Act and environmental protection. ➡ To stay up to date with the latest developments in AI policy & regulation and learn more about the AI Act, join 29,000+ people who subscribe to my weekly newsletter (link below). ➡ To learn more about the AI Act, join the AI, Tech & Privacy Academy's AI Act Bootcamp (link below). #AIRegulation #AIGovernance #AIAct #environmentalprotection

🚨 AI Governance IS HIRING! Below are 10 positions posted IN THE LAST FEW DAYS. Bookmark, share & be an early applicant: 1. Booking.com (Netherlands) - Group Data & AI Governance Manager.  ➡️ Apply: https://lnkd.in/eZMNUmgp 2. RBC (Canada) - Associate/VP, AI Governance  ➡️ Apply: https://lnkd.in/e3v-rwkq 3. AstraZeneca (Sweden) - Head of Data & AI Governance and Policy, R&D ➡️ Apply: https://lnkd.in/eRFJ_jK6 4. CGI (US) AI Governance Specialist ➡️ Apply: https://lnkd.in/eE6RVhBE 5. Swiss Re (Slovakia) - Data & AI Governance Architect ➡️ Apply: https://lnkd.in/eeu3gqNH 6. Venquis (Germany) - Senior Consultant - Generative AI Governance ➡️ Apply: https://lnkd.in/edtmsVkA 7. Jack Henry (US) - Corporate Counsel - Product and AI  ➡️ Apply: https://lnkd.in/ee_9P-Cb 8. Google (US) - Technical Program Manager, AI Governance Infrastructure ➡️ Apply: https://lnkd.in/eV3ZHxJq 9. First American (US) - Director of AI Governance and Information Security ➡️ Apply: https://lnkd.in/e6p-AJbf 10. GEICO (US) - Director, Data, Model & AI Governance  ➡️ Apply: https://lnkd.in/emp2rF4y ♻️ Share this post and help your network stay up to date with the latest AI Governance job opportunities. - 🚨 Additional information: ➡ If you are transitioning to AI governance or focusing on upskilling to land your dream job in AI, check out the AI, Tech & Privacy Academy's 4-week Bootcamps (upcoming cohorts starting NEXT WEEK). More than 800 people have joined our AI programs - don't miss them! ➡ For additional AI governance and privacy job opportunities, join thousands of people who receive the AI, Tech & Privacy Academy's weekly JOB ALERTS (link below). ➡ To stay up to date with the latest developments in AI policy & regulation and receive my insights, join 29,000+ people who subscribe to Luiza's Newsletter (link below). Good luck! #AIGovernance #AIRegulation #AIjobs #JobSearch #LinkedINJobs

🚨 BREAKING: The EU Commission announces that X is likely breaching the Digital Services Act. Elon Musk then accuses the EU of attempting to CENSOR speech. Here's what happened: ➡ Today, the EU Commission informed X of its preliminary view that the company is in breach of the Digital Services Act (DSA) regarding dark patterns, advertising transparency, and data access for researchers. ➡ According to the official press release (link below), these are the main issues that have motivated the preliminary findings: "➵ First, X designs and operates its interface for the 'verified accounts' with the 'Blue checkmark' in a way that does not correspond to industry practice and deceives users. Since anyone can subscribe to obtain such a 'verified' status, it negatively affects users' ability to make free and informed decisions about the authenticity of the accounts and the content they interact with. There is evidence of motivated malicious actors abusing the 'verified account' to deceive users. ➵ Second, X does not comply with the required transparency on advertising, as it does not provide a searchable and reliable advertisement repository, but instead put in place design features and access barriers that make the repository unfit for its transparency purpose towards users. In particular, the design does not allow for the required supervision and research into emerging risks brought about by the distribution of advertising online. ➵ Third, X fails to provide access to its public data to researchers in line with the conditions set out in the DSA. In particular, X prohibits eligible researchers from independently accessing its public data, such as by scraping, as stated in its terms of service. In addition, X's process to grant eligible researchers access to its application programming interface (API) appears to dissuade researchers from carrying out their research projects or leave them with no other choice than to pay disproportionally high fees." ➡ X can now defend itself. If the EU decides that X is breaching the DSA, the fine can be up to 6% of X's total worldwide annual turnover. ➡ In response to the EU Commission's announcement, Elon Musk posted: "The European Commission offered 𝕏 an illegal secret deal: if we quietly censored speech without telling anyone, they would not fine us. The other platforms accepted that deal. 𝕏 did not." ➡ It will be interesting to follow this legal battle - one more for Musk's collection. ➡ To stay up to date with the latest developments in AI & Tech policy & regulation, join 28,500 people who subscribe to my weekly newsletter (link below). #DigitalServicesAct #DSA #X #darkpatterns #techregulation #ElonMusk

🚨 BREAKING: Today, July 12th, the EU AI Act is published in the Official Journal, and the enforcement countdown starts. Here's what you must add to your calendar: 📌 August 1st, 2024: The AI Act will enter into force; 📌 February 2025: Chapters I (general provisions) & II (prohibited AI systems) will apply; 📌 August 2025: Chapter III Section 4 (notifying authorities), Chapter V (general purpose AI models), Chapter VII (governance), Chapter XII (confidentiality and penalties), and Article 78 (confidentiality) will apply, except for Article 101 (fines for General Purpose AI providers); 📌 August 2026: the whole AI Act applies, except for Article 6(1) & corresponding obligations (one of the categories of high-risk AI systems); 📌 August 2027 - Article 6(1) & corresponding obligations apply. ➡ The EU AI Act's official description is: "Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 laying down harmonised rules on artificial intelligence and amending Regulations (EC) No 300/2008, (EU) No 167/2013, (EU) No 168/2013, (EU) 2018/858, (EU) 2018/1139 and (EU) 2019/2144 and Directives 2014/90/EU, (EU) 2016/797 and (EU) 2020/1828 (Artificial Intelligence Act)Text with EEA relevance." ➡ Below is the official link to the law: don't forget to bookmark it. ➡ The AI, Tech & Privacy Academy's summer cohort of the EU AI Act Bootcamp starts NEXT WEEK. If you want to master the AI Act's main concepts and rules and get ready for regulatory challenges in AI, you can't miss it (link below). ➡ To stay up to date with the latest developments in AI policy & regulation, join 28,500+ people who subscribe to my weekly newsletter. #EUAIAct #AIAct #AIregulation #AIgovernance #AIpolicy #AILaw

🚨 [AI REGULATION] The definition and societal impacts of open foundation models should be at the core of AI policy & regulation discussions. Why? In many jurisdictions, there are more lenient rules for these AI models. Read this: ➡ The EU AI Act, for example, in its Article 53, which covers General-Purpose AI Models, establishes that: "2. The obligations set out in paragraph 1, points (a) and (b), shall not apply to providers of AI models that are released under a free and open-source licence that allows for the access, usage, modification, and distribution of the model, and whose parameters, including the weights, the information on the model architecture, and the information on model usage, are made publicly available. This exception shall not apply to general-purpose AI models with systemic risks." ➡ In the US, the Federal Trade Commission has recently published an article covering open-weights foundation models, highlighting their potential benefits (including enabling greater innovation, driving competition, improving consumer choice, and reducing costs) and possible risks to consumers when compared to centralized closed models (link to the article below). ➡ Given the high stakes and the fact that regulatory authorities will look at these AI models differently, both their definition and their societal impacts - especially in comparison to closed ones - must be closely scrutinized. ➡ In this context, the paper "On the Societal Impact of Open Foundation Models" by Sayash Kapoor, Rishi Bommasani, Kevin Klyman, Shayne Longpre, Ashwin Ramaswami for State Senate, Peter Cihon, Aspen Hopkins, Kevin Bankston, Stella Biderman, Miranda Bogen, Dr. Rumman Chowdhury, Alex Engler, Peter Henderson, Yacine Jernite, Seth Lazar, Stefano Maffulli, Alondra Nelson, Joelle Pineau, Aviya Skowron, Dawn Song, Victor Storchan, Daniel Zhang, Daniel E. Ho, Percy Liang & Arvind Narayanan is a must-read for everyone in AI, especially those focused on regulation and policymaking (link below). ➡ The paper discusses the distinctive properties of open foundation models, their benefits, a risk assessment framework to evaluate risks and threats, and recommendations and calls to action. ➡ AI liability is still an open topic in many jurisdictions worldwide. At this point, it is extremely important that policymakers and regulators invest time and effort in understanding the different risk profiles, the best way to regulate them, and who is responsible for the harm after it happens. ➡ Find all relevant links below. ➡ To stay up to date with the latest developments in AI policy & regulation, join 28,500 people who subscribe to my weekly newsletter (link below). #AI #FoundationModels #OpenFoundationModels #AIRegulation #AIgovernance #AIpolicy

🚨 [AI REGULATION] The Federal Trade Commission has just posted about Open-Weight Foundation Models, and it's a must-read for everyone in AI. Quotes: "Despite the nascent terminology, open-weights models, like OSS, have the potential to be a positive force for innovation and competition. Building on the traditions of open-source software, open-weights models can, in principle, be incorporated into a product without going through a procurement process and can be tweaked and fine-tuned to suit the needs of the user (e.g., a product developer researcher, or scientist), providing flexibility to those who want to use advanced models without forcing them to train a model from scratch. Open-weights models can further enable competition by offering portability between different hosting providers." - "In addition to recognizing the potential benefits, it is also important to be attuned to restrictions associated with some models claimed as “open.” For example, a model developer may make the model’s weights available, but only under licensing terms that are so restrictive as to effectively prevent its use in the marketplace – such as by prohibiting commercial use. Models with such licenses may be useful for research but are unlikely to directly promote competition. Other license restrictions could have a similar dampening effect on open-weights models’ benefits to competition." - "But open-weights models also pose additional risks to consumers over centralized closed models. The lowered costs and barriers to retraining and redistributing models extends not only to well-meaning companies but also to malicious actors scaling up spam, scams, and other harmful uses." ➡ Link to the article below. ➡ To stay up to date with the latest developments in AI policy & regulation, join 28,500 people who subscribe to my newsletter (link below). #AI #AIRegulation #AIpolicy #FTC #FoundationModels #AIGovernance

🚨 BREAKING: 28,500+ people have just received the 111th edition of my newsletter on AI policy & regulation, and you CAN'T MISS it! Here's what you'll find: ➵ Tracking AI legislation ➵ Meta is having a hard time in the EU ➵ Meta is also having a hard time in Brazil ➵ Generative AI patents ➵ Course AI Assistants: Legal & Ethical Implications ➵ AI & Privacy paper ➵ AI Governance is HIRING ➵ Insights on AI governance, compliance & regulation ➵ Upcoming AI training programs ➡ READ & SUBSCRIBE: https://lnkd.in/guJRMWC4 ➡ If you think your network will benefit from the newsletter, consider sharing. Thank you! #AI #AIGovernance #AIregulation #AILegislation #GenerativeAI #privacy