What's new arround internet
| Src | Date (GMT) | Titre | Description | Tags | Stories | Notes |
 |
2024-03-07 11:00:00 | Sécuriser l'IA Securing AI (lien direct) |
With the proliferation of AI/ML enabled technologies to deliver business value, the need to protect data privacy and secure AI/ML applications from security risks is paramount. An AI governance framework model like the NIST AI RMF to enable business innovation and manage risk is just as important as adopting guidelines to secure AI. Responsible AI starts with securing AI by design and securing AI with Zero Trust architecture principles. Vulnerabilities in ChatGPT A recent discovered vulnerability found in version gpt-3.5-turbo exposed identifiable information. The vulnerability was reported in the news late November 2023. By repeating a particular word continuously to the chatbot it triggered the vulnerability. A group of security researchers with Google DeepMind, Cornell University, CMU, UC Berkeley, ETH Zurich, and the University of Washington studied the “extractable memorization” of training data that an adversary can extract by querying a ML model without prior knowledge of the training dataset. The researchers’ report show an adversary can extract gigabytes of training data from open-source language models. In the vulnerability testing, a new developed divergence attack on the aligned ChatGPT caused the model to emit training data 150 times higher. Findings show larger and more capable LLMs are more vulnerable to data extraction attacks, emitting more memorized training data as the volume gets larger. While similar attacks have been documented with unaligned models, the new ChatGPT vulnerability exposed a successful attack on LLM models typically built with strict guardrails found in aligned models. This raises questions about best practices and methods in how AI systems could better secure LLM models, build training data that is reliable and trustworthy, and protect privacy. U.S. and UK’s Bilateral cybersecurity effort on securing AI The US Cybersecurity Infrastructure and Security Agency (CISA) and UK’s National Cyber Security Center (NCSC) in cooperation with 21 agencies and ministries from 18 other countries are supporting the first global guidelines for AI security. The new UK-led guidelines for securing AI as part of the U.S. and UK’s bilateral cybersecurity effort was announced at the end of November 2023. The pledge is an acknowledgement of AI risk by nation leaders and government agencies worldwide and is the beginning of international collaboration to ensure the safety and security of AI by design. The Department of Homeland Security (DHS) CISA and UK NCSC joint guidelines for Secure AI system Development aims to ensure cybersecurity decisions are embedded at every stage of the AI development lifecycle from the start and throughout, and not as an afterthought. Securing AI by design Securing AI by design is a key approach to mitigate cybersecurity risks and other vulnerabilities in AI systems. Ensuring the entire AI system development lifecycle process is secure from design to development, deployment, and operations and maintenance is critical to an organization realizing its full benefits. The guidelines documented in the Guidelines for Secure AI System Development aligns closely to software development life cycle practices defined in the NSCS’s Secure development and deployment guidance and the National Institute of Standards and Technology (NIST) Secure Software Development Framework (SSDF). The 4 pillars that embody the Guidelines for Secure AI System Development offers guidance for AI providers of any systems whether newly created from the ground up or built on top of tools and services provided from | Tool Vulnerability Threat Mobile Medical Cloud Technical | ChatGPT | ★★ |
 |
2024-01-31 10:30:00 | ESET Research Podcast: Chatgpt, The Moveit Hack et Pandora ESET Research Podcast: ChatGPT, the MOVEit hack, and Pandora (lien direct) |
Un chatbot AI Kindle par inadvertance un boom de la cybercriminalité, des bandits de ransomware pluncent des organisations sans déploiement
An AI chatbot inadvertently kindles a cybercrime boom, ransomware bandits plunder organizations without deploying ransomware, and a new botnet enslaves Android TV boxes |
Ransomware Hack Mobile | ChatGPT | ★★★ |
 |
2024-01-05 21:36:17 | Les utilisateurs d'Android pourraient bientôt remplacer Google Assistant par Chatgpt Android users could soon replace Google Assistant with ChatGPT (lien direct) |
L'application Android Chatgpt travaille sur la prise en charge des API assistantes d'Android \\.
The Android ChatGPT app is working on support for Android\'s assistant APIs. |
Mobile | ChatGPT | ★★ |
 |
2023-12-22 22:47:44 | Rapport de menace ESET: abus de nom de chatppt, Lumma Steal Maleware augmente, la prévalence de Spyware \\ Android Spinok SDK \\ ESET Threat Report: ChatGPT Name Abuses, Lumma Stealer Malware Increases, Android SpinOk SDK Spyware\\'s Prevalence (lien direct) |
Des conseils d'atténuation des risques sont fournis pour chacune de ces menaces de cybersécurité.
Risk mitigation tips are provided for each of these cybersecurity threats. |
Malware Threat Mobile | ChatGPT | ★★★ |
 |
2023-11-28 23:05:04 | Prédictions 2024 de Proofpoint \\: Brace for Impact Proofpoint\\'s 2024 Predictions: Brace for Impact (lien direct) |
In the ever-evolving landscape of cybersecurity, defenders find themselves navigating yet another challenging year. Threat actors persistently refine their tactics, techniques, and procedures (TTPs), showcasing adaptability and the rapid iteration of novel and complex attack chains. At the heart of this evolution lies a crucial shift: threat actors now prioritize identity over technology. While the specifics of TTPs and the targeted technology may change, one constant remains: humans and their identities are the most targeted links in the attack chain. Recent instances of supply chain attacks exemplify this shift, illustrating how adversaries have pivoted from exploiting software vulnerabilities to targeting human vulnerabilities through social engineering and phishing. Notably, the innovative use of generative AI, especially its ability to improve phishing emails, exemplifies a shift towards manipulating human behavior rather than exploiting technological weaknesses. As we reflect on 2023, it becomes evident that cyber threat actors possess the capabilities and resources to adapt their tactics in response to increased security measures such as multi-factor authentication (MFA). Looking ahead to 2024, the trend suggests that threats will persistently revolve around humans, compelling defenders to take a different approach to breaking the attack chain. So, what\'s on the horizon? The experts at Proofpoint provide insightful predictions for the next 12 months, shedding light on what security teams might encounter and the implications of these trends. 1. Cyber Heists: Casinos are Just the Tip of the Iceberg Cyber criminals are increasingly targeting digital supply chain vendors, with a heightened focus on security and identity providers. Aggressive social engineering tactics, including phishing campaigns, are becoming more prevalent. The Scattered Spider group, responsible for ransomware attacks on Las Vegas casinos, showcases the sophistication of these tactics. Phishing help desk employees for login credentials and bypassing MFA through phishing one-time password (OTP) codes are becoming standard practices. These tactics have extended to supply chain attacks, compromising identity provider (IDP) vendors to access valuable customer information. The forecast for 2024 includes the replication and widespread adoption of such aggressive social engineering tactics, broadening the scope of initial compromise attempts beyond the traditional edge device and file transfer appliances. 2. Generative AI: The Double-Edged Sword The explosive growth of generative AI tools like ChatGPT, FraudGPT and WormGPT bring both promise and peril, but the sky is not falling as far as cybersecurity is concerned. While large language models took the stage, the fear of misuse prompted the U.S. president to issue an executive order in October 2023. At the moment, threat actors are making bank doing other things. Why bother reinventing the model when it\'s working just fine? But they\'ll morph their TTPs when detection starts to improve in those areas. On the flip side, more vendors will start injecting AI and large language models into their products and processes to boost their security offerings. Across the globe, privacy watchdogs and customers alike will demand responsible AI policies from technology companies, which means we\'ll start seeing statements being published about responsible AI policies. Expect both spectacular failures and responsible AI policies to emerge. 3. Mobile Device Phishing: The Rise of Omni-Channel Tactics take Centre Stage A notable trend for 2023 was the dramatic increase in mobile device phishing and we expect this threat to rise even more in 2024. Threat actors are strategically redirecting victims to mobile interactions, exploiting the vulnerabilities inherent in mobile platforms. Conversational abuse, including conversational smishing, has experienced exponential growth. Multi-touch campaigns aim to lure users away from desktops to mobile devices, utilizing tactics like QR codes and fraudulent voice calls | Ransomware Malware Tool Vulnerability Threat Mobile Prediction Prediction | ChatGPT ChatGPT | ★★★ |
 |
2023-11-22 18:39:21 | «Chatgpt with Voice» s'ouvre à tout le monde sur iOS et Android “ChatGPT with voice” opens up to everyone on iOS and Android (lien direct) |
Tous les utilisateurs d'Android et iOS peuvent bientôt appuyer sur une icône de casque et commencer à discuter.
All Android and iOS users can soon tap a headphone icon and start chatting. |
Mobile | ChatGPT | ★★ |
1
We have: 6 articles.
We have: 6 articles.
To see everything:
Our RSS (filtrered)
