The hottest Privacy Substack posts right now

And their main takeaways
Category
Top U.S. Politics Topics

Must Learn AI Security Part 14: Inference Attacks Against AI

Rod’s Blog • 39 implied HN points • 18 Sep 23
🕹 Technology AI Security Privacy Data Protection Security Measures
  1. An inference attack against AI involves gaining private information from a system by analyzing its outputs and other available data.
  2. There are two main types of inference attacks: model inversion attacks aim to reconstruct input data, while membership inference attacks try to determine if specific data points were part of the training dataset.
  3. To mitigate inference attacks, techniques like differential privacy, federated learning, secure multi-party computation, data obfuscation, access control, and regular model updates can be used.

Must Learn AI Security Part 7: Membership Inference Attacks Against AI

Rod’s Blog • 39 implied HN points • 24 Aug 23
🕹 Technology AI Security Privacy Data Protection Machine Learning
  1. Membership Inference Attacks against AI involve attackers trying to determine if a specific data point was part of a machine learning model's training dataset by analyzing the model's outputs.
  2. These attacks occur in steps like data collection, model access, creating shadow models, analyzing model outputs, and making inferences based on the analysis.
  3. The consequences of successful Membership Inference Attacks include privacy violations, data leakage, regulatory risks, trust erosion, and hindrance to data sharing in AI projects.

Localisation of Gen AI in EU

Sector 6 | The Newsletter of AIM • 39 implied HN points • 01 Sep 23
🕹 Technology AI Data Privacy Regulation Innovation
  1. The EU has strict data protection laws that make it hard for AI tools like ChatGPT to work there. Companies have to follow these rules carefully.
  2. European lawmakers are banning certain AI technologies, like biometric surveillance and predictive policing. This is changing how AI innovations happen in Europe.
  3. A French company called Mistral AI recently raised a lot of money, even though they haven't launched a product yet. Their team has a lot of experience in developing advanced AI models.

TechLetters #121 - Cyberwar in Ukraine. 0day exploits to smartphones via apps. AI-automation coming to address cybersecurity industry. Cookie blocking standardisation. VULKAN cyberwar services.

Lukasz Olejnik on Cyber, Privacy and Tech Policy Critique • 39 implied HN points • 03 Apr 23
🕹 Technology Security Privacy Technology Policy Other
  1. Smartphone app exploited 0day vulnerability affecting millions of devices
  2. Microsoft integrating ChatGPT for automating cybersecurity tasks
  3. Web security proposal aims to standardize cross-browser cookie blocking

TechLetters #126 - Banning multi-sim devices. Russian stealthy espionage combine harvester analysed and made public. Putting AI-generated labels on AI-generated content.

Lukasz Olejnik on Cyber, Privacy and Tech Policy Critique • 39 implied HN points • 15 May 23
🕹 Technology Security Privacy Technology Policy Other
  1. UK is considering banning multi-sim devices to prevent fraud
  2. US reveals Russian cyber espionage tool used by FSB for information theft
  3. Google to label AI-generated images to comply with EU regulations
Get a weekly roundup of the best Substack posts, by hacker news affinity:

Language Models: Size Matters

Fully Distributed by Ori Eldarov • 39 implied HN points • 30 Mar 23
🕹 Technology Artificial Intelligence Machine Learning Open Source Privacy
  1. The trend towards large language models (LLMs) may not be the best approach due to high training costs and lack of optimization.
  2. Research shows that smaller language models can perform better through fine-tuning with human feedback, offering cost-efficiency and hyper-personalization.
  3. The future may see a mix of ultra-large proprietary models and small open-source models, working together to advance artificial intelligence.

Must Learn AI Security Part 6: Model Inversion Attacks Against AI

Rod’s Blog • 39 implied HN points • 23 Aug 23
🕹 Technology AI Security Privacy Data Breach Cybersecurity Machine Learning
  1. A Model Inversion attack against AI involves reconstructing training data by only having access to the model's output, posing risks to data privacy.
  2. There are two main types of Model Inversion attacks: black-box attack and white-box attack, differing in the level of access the attacker has to the AI model.
  3. Model Inversion attacks can have severe consequences like privacy violation, identity theft, loss of trust, legal issues, and misuse of sensitive information, emphasizing the need for robust security measures.

Meta Blocker

Technically Optimistic • 19 implied HN points • 02 Feb 24
🇺🇸 U.S. Politics Social media Legislation Child safety Privacy Law enforcement
  1. Reports of online child abuse have significantly increased over the years, showcasing a concerning trend that AI technology might exacerbate.
  2. Tech companies, like Snapchat and X, are starting to agree to support legislative acts like KOSA to enhance online protections for children, signaling a potential positive shift in the industry.
  3. There is a growing call for tougher legislation, transparency, and accountability in the tech sector to protect users, especially children, from harmful content online.

Brief: How Microsoft Secures Its Copilots

Rod’s Blog • 19 implied HN points • 01 Feb 24
🕹 Technology AI Privacy Security Data Protection Machine Learning
  1. Microsoft's Copilot for Microsoft 365 adheres to strict data privacy and security regulations like GDPR, ensuring organizational data confidentiality.
  2. The Copilot system integrates large language models with Microsoft Graph and 365 apps, maintaining enterprise-level data protection during processing.
  3. By utilizing the Azure OpenAI Service controlled by Microsoft, Copilot ensures that business data is not used to train models, offering organizations control over their data processing.

Staying Safe in a World Full of AI Integrated Devices

Rod’s Blog • 19 implied HN points • 31 Jan 24
🕹 Technology Privacy Security Well-being
  1. AI can pose risks to privacy through data collection without consent; protect your privacy with strong passwords and limit AI features' access.
  2. AI can threaten security through sophisticated attacks like deepfakes; protect your security with regular updates, antivirus software, and verifying content sources.
  3. AI can impact well-being by increasing stress and reducing social skills; protect your well-being by setting boundaries, balancing online and offline activities, and maintaining social connections.

TechLetters #160 Resurrection of dead dictators using generative AI, for political gain; AI cyber operations; Last warning about quantum cryptography; Privacy Sandbox coming to iOS/iPhones

Lukasz Olejnik on Cyber, Privacy and Tech Policy Critique • 19 implied HN points • 29 Jan 24
🕹 Technology Security Privacy Technology Policy
  1. AI can resurrect dead politicians, prompting the need for laws against misleading content.
  2. iOS introduces Stolen Device Protection and warns about AI's impact on cyber operations.
  3. Caution advised on quantum cryptography and factors influencing ransom payments.

Reaching the enchanted forest

Superficial Intelligence • 18 HN points • 19 Jun 24
🕹 Technology AI Design Innovation Human-Computer Privacy
  1. Nature used to feel more alive and connected, and it inspired our stories and spirituality. Many people believe that the natural world has a spirit, making it important to recognize its role in our lives.
  2. Modern living spaces are often designed for practicality but can feel boring and unresponsive. We can use technology to make our environments more interactive and supportive, like a helpful friend.
  3. Embedded AI offers a way to change our tools and devices to understand our needs better. This can help us regain a sense of ease and connection, making our daily lives less dependent on screens and more engaging.

The Rise of Surveillance Bureaucracy

The Digital Anthropologist • 59 implied HN points • 01 Jan 23
🕹 Technology Surveillance Bureaucracy Digital technology Privacy Ethics
  1. Surveillance bureaucracy uses digital technologies like sensors, cameras, and monitoring software to improve efficiency and compliance in both government and business.
  2. The rise of surveillance bureaucracy has unintended consequences, leading to resistance and workarounds by surveilled individuals in various industries.
  3. Challenges from non-profit organizations and legal battles are emerging to address the ethical implications of digital surveillance technologies and the need for privacy protections.

Practical Reinforcement Learning and Differential Privacy

Gradient Flow • 119 implied HN points • 17 Feb 22
🕹 Technology Machine Learning Data science Infrastructure Privacy
  1. The ratio of data scientists to data engineers varies based on factors like tools, infrastructure, and use cases, with no set ideal ratio.
  2. Interesting developments include a new podcast discussing machine learning infrastructure at Netflix, imperceptible NLP attacks, and evolving data science training programs.
  3. Exciting tools and updates in the data and machine learning space, like practical reinforcement learning applications, scalable differential privacy for Python developers, and the Orbit version 1.1 for Bayesian time-series analysis.

Want to help me upgrade digital identity?

Steve Kirsch's newsletter • 4 implied HN points • 01 Feb 25
🕹 Technology Digital Identity Cryptography Cloud Security Privacy
  1. Usernames and passwords are outdated. A new method of digital identity would make online security simpler and safer.
  2. You can manage your identity without needing to remember complicated passwords or codes. Just an alias is all you need.
  3. There’s a need for experts in cryptography and cloud security to help create a more secure and self-sovereign identity system.

Proposal for improving the global online discourse through personalised comment ordering on all websites

Engineering Ideas • 19 implied HN points • 07 Dec 23
🕹 Technology Social media AI Data Blockchain Privacy
  1. Social media promotes tribalism and polarization, making it hard to find rational critique in comments.
  2. A proposed solution involves personalized comment ordering based on user reactions and models.
  3. Compensating users for reading and voting on comments with a token system could help combat spam and manipulation.

TechLetters #116 - cyberwar in Ukraine; TikTok gets a hit in Europe; 2-year-old backdoor activated in cyberwar in Ukraine; #GDPR reform imminent; Privacy Sandbox field-tested; geopolitical censorship

Lukasz Olejnik on Cyber, Privacy and Tech Policy Critique • 19 implied HN points • 27 Feb 23
🕹 Technology Security Privacy Technology Policy Other
  1. Analysis of cyberwar in Ukraine with details of satellite internet provider KA-SAT compromise
  2. EU institutions prohibit TikTok use by employees for cybersecurity reasons
  3. European Commission proposes GDPR reform for more coherent enforcement rules

Thiel-backed startup began contracting on "off-grid" facial recognition with U.S. Air Force

All-Source Intelligence Fusion • 40 implied HN points • 21 Aug 23
🕹 Technology Facial Recognition Startups Military Tech Privacy Surveillance
  1. The Thiel-backed startup goTenna is working with the U.S. Air Force on off-grid facial recognition.
  2. Clearview AI also had similar contracts with the Air Force for facial recognition technology.
  3. goTenna's expansion into military facial recognition is attracting more attention and scrutiny.

TechLetters #136 Reviewing for security and privacy. ChatGPT for cybercrime. Privacy Sandbox, AI, sending e-mails a problem in 2023.

Lukasz Olejnik on Cyber, Privacy and Tech Policy Critique • 19 implied HN points • 24 Jul 23
🕹 Technology Security Privacy Technology Policy Other
  1. Sending military emails requires caution to avoid leaking personal data.
  2. ChatGPT is used for illegal activities like phishing, with no ethical restrictions.
  3. The launch of the Privacy Sandbox by Google/Chrome aims to improve privacy in ad infrastructure.

TechLetters #127 - Russian citizen sentenced for participation in IT Army of Ukraine. Google to identify A used in election disinformation. China bans US-made chips (Micron). Cookies out in 2024.

Lukasz Olejnik on Cyber, Privacy and Tech Policy Critique • 19 implied HN points • 22 May 23
🕹 Technology Security Privacy Technology Policy
  1. Russian citizen sentenced to 3 years for participating in IT Army of Ukraine
  2. Google aims to identify and block election disinformation campaigns using AI
  3. China bans the use of US-made Micron chips citing security risks

🥟 Chao-Down #78 Hackers use generative AI to create malware, Google introduces AI-powered privacy platform, ChatGPT begins giving therapy

Chaos Theory • 19 implied HN points • 04 May 23
🕹 Technology AI Privacy Malware Therapy Cybersecurity
  1. Hackers are using generative AI to create malware, making it harder for cybersecurity professionals to detect and defend against these attacks.
  2. Google introduces an AI-powered privacy platform called Checks.
  3. ChatGPT is starting to give therapy, potentially revolutionizing mental health care.

TechLetters #134 cyberattack on WATER TREATMENT. GDPR reform. Competition authorities enforcement GDPR. France considers social network shutdown during unrest. Avast collects/sells customer data, and?

Lukasz Olejnik on Cyber, Privacy and Tech Policy Critique • 19 implied HN points • 10 Jul 23
🕹 Technology Security Privacy Technology Policy Generative AI
  1. Cybersecurity incidents on critical infrastructure are a major concern globally
  2. Proposed GDPR reforms aim to enhance enforcement mechanisms and transparency
  3. Countries are implementing stricter regulations to protect data privacy and crack down on unethical data practices

TechLetters #132 GDPR fine for Criteo. Artificial Intelligence Act lobbying. In-game malware. processors leaking private data.

Lukasz Olejnik on Cyber, Privacy and Tech Policy Critique • 19 implied HN points • 26 Jun 23
🕹 Technology Security Privacy Technology Policy
  1. Malware was found in a Mario Bros game installer, which steals user data.
  2. Privacy leaks are occurring via processors, allowing websites to steal browsing history.
  3. Criteo received a €40M GDPR fine for not verifying if people had given consent for their data processing.

AImagine: Pioneering Hyperrealistic AI Image Generation

CodeLink’s Substack • 19 implied HN points • 18 May 23
🕹 Technology AI Image Generation DevOps Security Privacy
  1. AI technology is revolutionizing image generation and manipulation, offering new creative possibilities and demand
  2. AImagine app by CodeLink stands out for its hyperrealistic results and high level of customization in generating unique images
  3. Utilizing innovative technologies like the stable diffusion model, Flutter, and Python, AImagine offers a seamless user experience and efficient server-side processing

Personalization Uncovered: The Surprising Ways Hyper-Targeted Advertising Influences Our Thoughts, Choices, and Lives

⭐️Bob’s Newsletter • 19 implied HN points • 30 Mar 23
🕹 Technology Advertising Data Privacy Personalization
  1. Hyper-targeted advertising limits exposure to new ideas and unexpected discoveries.
  2. Personalized advertising may give a false sense of autonomy in our consumption choices.
  3. Repetitive hyper-targeted ads can diminish real emotional connections with brands and others.

Social Regression

Venture Prose • 239 implied HN points • 30 Dec 19
🕹 Technology Social media Privacy Information Technology Digital detox Personal Development
  1. Modernization can lead to excess and intellectual regression, causing a cycle of abuse and confinement in technology.
  2. Balancing digital consumption and real-life experiences can lead to a happier and more focused lifestyle.
  3. Cutting distractions and focusing on what truly matters can lead to a more fulfilled and purposeful life, while still enjoying the benefits of technology.

11 steps to keep Meta from stealing your data to train AI

Ladyparts • 2 HN points • 18 Jun 24
🕹 Technology Privacy Social media AI Data Protection
  1. You only have until June 26, 2024 to opt out of Meta using your data for AI training - act before the deadline.
  2. The process to opt out of Meta's data use can be confusing and intentionally complicated - it's important to follow the steps carefully.
  3. The steps provided include logging into Facebook on a laptop, navigating through privacy settings, and being persistent in filling out the opt-out form.

A tale of iCloud account recovery

Kamil’s Substack • 3 HN points • 14 May 24
🕹 Technology Security Privacy Authentication Payment processing Digital Identity
  1. During iCloud account recovery, you may be asked for credit card details that are actually verified by running a charge, causing issues even with correct information.
  2. Securing your own email account can involve user-controlled methods like two-factor authentication with a physical token, whereas iCloud's security measures are more restrictive and dictated by the service provider.
  3. Recovering an iCloud account might involve providing credit card details, which are tested by running a transaction, leading to potential issues if the card details change.

Watermark the world

escape the algorithm • 59 implied HN points • 18 Mar 22
🕹 Technology Internet Ownership Privacy Ethics Data
  1. Google Street View is made up of images from various sources, including everyday people, blended together to create a seamless representation of the world.
  2. Watermarks added to Google Street View images are intentional, potentially highlighting the hidden labor behind the scans or symbolizing colonialism by claiming ownership of public spaces.
  3. Question arises: Is the act of watermarking in Google Street View a way to show presence or a form of staking a claim on territory?

Threat Model

Reboot • 21 implied HN points • 18 Nov 23
🕹 Technology Privacy Security Internet Surveillance Digital ethics
  1. In the age of the internet, it's challenging to disappear in America due to the digital footprint left by basic needs like addresses, bank cards, and online accounts.
  2. There is a small community of tech-savvy individuals focused on extreme privacy, using specialized tools like privacy-focused operating systems, encrypted communication, and anonymous browsing.
  3. Extreme privacy can be isolating and exhausting, leading to a constant balance between the desire for privacy and the inherent risks and limitations of disappearing from the public eye.

Người ta có thể xác định vị trí của blogger hay không?

Thái | Hacker | Kỹ sư tin tặc • 79 implied HN points • 27 Aug 21
🕹 Technology Blogging Online Security Privacy
  1. Encrypting blog traffic with HTTPS can make it difficult for someone to track or determine a blogger's location.
  2. Be cautious about clicking on links in comments to prevent revealing your IP address.
  3. Taking precautions like using incognito mode, avoiding excessive self-visits to your blog, and using secure DNS settings can help protect your identity and location online.

Lobster long-read #1: A short history of unlocking things

Design Lobster • 119 implied HN points • 12 Nov 20
🕹 Technology Privacy Security Internet Data Digital
  1. Locks have evolved over time, from simple mechanisms like holes in doors to more complex designs with pins and tumblers, highlighting the importance of privacy and security in history.
  2. The mental model of a lock, where a key unlocks a 'private' space, is now applied to digital privacy, but the reality is that we entrust our digital possessions to third parties online.
  3. An alternative paradigm for online privacy involves incorporating detection mechanisms, like Apple's iOS alerts, to make visible the handling of our digital data by third parties and promote transparency.

Chiếc iPad trong vụ án Nguyễn Đức Chung

Thái | Hacker | Kỹ sư tin tặc • 59 implied HN points • 31 Dec 21
🌍 World Politics Legal Technology Privacy Government Corruption
  1. The importance of digital evidence, like an iPad, in legal cases and the challenges of accessing and verifying such data.
  2. The complex legal implications around passwords, access to personal devices, and email privacy in investigations.
  3. The limitations of access to Gmail and iPad data by law enforcement, highlighting the need for strong security measures to protect personal information.