LukaszOlejnik

LukaszOlejnik, 22 hours ago to random

Breakthrough judgment of European Court of Human Rights: states to provide judicial oversight & independent monitoring of surveillance. States must notify individuals subjected to surveillance. Evidence from illegal surveillance should not be used https://hudoc.echr.coe.int/app/conversion/pdf/?library=ECHR&id=003-7957259-11092429&filename=Judgment%20Pietrzak%20and%20Bychawska-Siniarska%20and%20Others%20v.%20Poland%20-%20complaints%20about%20Polish%20legislation%20on%20secret%20surveillance.pdf

LukaszOlejnik, 18 hours ago to random

Israel allegedly conducted cyberespionage operations on International Criminal Court (ICC) for 9 years. Allegedly to intimidate the ICC and disrupt investigations into potential Israeli war crimes in Gaza. https://www.theguardian.com/world/article/2024/may/28/spying-hacking-intimidation-israel-war-icc-exposed

LukaszOlejnik, 2 days ago to random

My newsletter #TechLetters out (unsure when the next one would be released). Data protection of TLS Session Tickets. Cyberattacks vs water utilities. EU wants to solve disinformation. Celebrities continue getting AI transparency immediately. https://open.substack.com/pub/techletters/p/techletters-175-data-protection-of

LukaszOlejnik, 2 days ago to random

Chinese army field-testing an assault rifle walking robot, and a quadcopter. Used to clear the area. International Humanitarian Law will be increasingly challenged. Not tomorrow but today.

video/mp4

LukaszOlejnik, 2 days ago to random

The AI Overview in the core, critical, Google search product underwent some form of risk assessment. So they knew about these issues, deciding to accept the risk. No sane person would jump off a bridge because a search engine advised it, right? As Google said, such quirky advice is given in a small number of cases. Perhaps those few individuals who might follow the strange advice (glue, bridge, rocks, etc.) were deemed an acceptable risk, a business decision that someone had to approve?

LukaszOlejnik, 3 days ago to random

Failed gurus claim that their prognostication was essentially correct although it might appear otherwise to us; or they tell us that the date at which their thesis will be confirmed has not yet arrived.

LukaszOlejnik, 5 days ago to random

Statement of 16 EU Member States about fighting disinformation and propaganda. „essential to strengthen ability to detect foreign information manipulation and interference to swiftly unveil them, to break their virality”. Defensive information activities to be extended to candidates. https://www.gov.pl/web/diplomacy/statement-by-the-ministers-for-european-affairs-of-france-germany-poland-austria-bulgaria-croatia-czechia-denmark-greece-italy-latvia-luxembourg-portugal-romania-slovenia-spain2

LukaszOlejnik, 6 days ago to random

China urging the country's domestic car manufacturers (SAIC, BYD, Dongfeng, GAC, etc) to use 25% Chinese chips by 2025. They are building a local manufacturing industry. The key is to build local demand. Here's how it's done. https://asia.nikkei.com/Business/Automobiles/China-asks-carmakers-to-use-up-to-25-local-chips-by-2025

LukaszOlejnik, 13 days ago to random

My comments in TechCrunch. Google/Android AI feature to detect scams-calls is useful. It's client-side so privacy guaranteed. However, once client-side scanning and control is is baked into software infrastructure, nobody knows where it may lead. Censorship? Technology control of societies? Let's just be careful here. https://techcrunch.com/2024/05/15/googles-call-scanning-ai-could-dial-up-censorship-by-default-privacy-experts-warn/

LukaszOlejnik, 6 days ago to random

China releases an AI chatbot trained on the thoughts of President Xi Jinping, the political philosophy “Xi Jinping Thought on Socialism with Chinese Characteristics for a New Era”, as well as other official literature. Competition for OpenAI? https://www.ft.com/content/43378c6e-664b-4885-a255-31325d632ee9

LukaszOlejnik, 6 days ago to random

I sure hope that Canon's website wasn't delivering backdoored drivers today. These days you never know.

LukaszOlejnik, 6 days ago to random

Fingerprinting web browsers using CSS features to infer system characteristics like the list of supported fonts. https://par.nsf.gov/servlets/purl/10425906

image/png
image/png
image/png

LukaszOlejnik, 6 days ago to random

My data protection assessment of TLS Session Tickets. While they are a unique identifier, all is OK. No consent needed. Another conclusion would mean that the data protection rules are inherently inadequate and harmful. #GDPR #ePrivacy #CCPA #APRA https://blog.lukaszolejnik.com/tls-session-tickets-and-data-protection-gdpr-eprivacy/

LukaszOlejnik, 8 days ago to random

Scarlet Johansson's complaint on OpenAI using 'voice similar to her' in ChatGPT voice assistant. "demanding that OpenAI disclose how it developed...". . https://www.npr.org/2024/05/20/1252495087/openai-pulls-ai-voice-that-was-compared-to-scarlett-johansson-in-the-movie-her

Celebrities have it easier. In your or my case it's harder. Hence: my #GDPR complaint: https://blog.lukaszolejnik.com/ai-llms-gdpr-complaint-and-human-dignity/

LukaszOlejnik, 8 days ago to random

Scarlet Johansson's complaint on OpenAI using 'voice similar to her' in the latest release of ChatGPT voice assistant. "demanding that OpenAI disclose how it developed an AI personal assistant voice" https://www.npr.org/2024/05/20/1252495087/openai-pulls-ai-voice-that-was-compared-to-scarlett-johansson-in-the-movie-her

LukaszOlejnik, 8 days ago to random

Warning about cyberattacks against water utilities. Immediate actions needed to protect drinking water. Impacts may include interruptions to water treatment and storage; damage to pumps and valves; and alteration of chemical levels to hazardous amounts. https://www.epa.gov/enforcement/enforcement-alert-drinking-water-systems-address-cybersecurity-vulnerabilities

LukaszOlejnik, 8 days ago to random

On October 22 in Bern I'm speaking at Swiss Cyber Storm. Topic to be announced later

LukaszOlejnik, 11 days ago to random

Cyber threat actor directing operations at AI researchers and experts (in academia, industry, government). Suspicion on Chinese(?) group. The goal is data theft. https://www.proofpoint.com/us/blog/threat-insight/security-brief-artificial-sweetener-sugargh0st-rat-used-target-american

LukaszOlejnik, 13 days ago to random

Google unveiled an LLM model that can monitor phone calls to detect suspected fraud. This technology indicates that capabilities to monitor calls, texts, and documents for undesirable content already exist or are being developed. Such a model could issue warnings, block actions, or report content, leading to technological modulation of social behavior. This poses a major threat to privacy and basic freedoms.

Read about designing technologies with values
https://blog.lukaszolejnik.com/designing-technologies-with-values-possibility-necessity-and-european-story/

LukaszOlejnik, 13 days ago to random

Writing a technical-data protection analysis about an element of core internet security. For fun I also discussed the topic with GPT4. Model's observations, reasoning and conclusion turned out to be ... completely wrong. Such a conclusion would make internet use in EU impossible.

LukaszOlejnik, 14 days ago to random

Google's I/O AI/LLM capabilities were much more impressive than the OpenAI one. Especially those parts devoted to helping people with disabilities. Actual useful contribution.

LukaszOlejnik, 14 days ago to random

Spelling AI doom fear no longer fashionable? Companies and investors pouring dollars to block strict safety rules on AI. A change from the AI debate dominated by well-funded lobby scare about the dangers of AI. https://www.politico.com/news/2024/05/12/ai-lobbyists-gain-upper-hand-washington-00157437

LukaszOlejnik, 14 days ago to random

President of USA is directing increases in tariffs across strategic sectors such as steel and aluminum, semiconductors, electric vehicles, batteries, critical minerals, solar cells, ship-to-shore cranes, and medical products. The tariff rate on semiconductors will increase from 25% to 50% by 2025.

LukaszOlejnik, 15 days ago to random

Chinese companies will soon be able to build humanoid robots on a massive scale. Today costing $14,000, will be even cheaper. They will be able to do more. I estimate that by 2030 there is no way to reliably equip them with weapons with significant effect https://twitter.com/xuxin_cheng/status/1789918278095151534/video/1

LukaszOlejnik, 15 days ago to random

I finished the first stage of editorial work on my upcoming book Propaganda. Sent to the Publisher. Now focusing on other tasks, and awaiting for the final stuff to come in June, and a release in October :) https://blog.lukaszolejnik.com/propaganda-my-book-on-information-security/