Research
Defenders Use Prompt Injections to Stop AI Attacks
WIRED AIhigh confidence
Why it matters
- →Context bombing offers a new defensive strategy against AI hacking agents.
- →It significantly reduces the success rate of AI-driven attacks.
- →This technique leverages the inherent vulnerabilities of prompt injections for security.
Researchers at Tracebit have developed a novel defensive technique called 'context bombing' to thwart AI hacking agents. By embedding prompt injections alongside sensitive data in AWS environments, they can cause large language models to shut down when encountering these prompts. Tests showed a significant reduction in the success rate of AI-driven attacks, with admin privilege escalation dropping from 57% to 5%. This innovative approach offers a new strategy for enhancing AI security defenses.
Read originalMore from WIRED AI
© WIRED AIMarket & Regulationmusic
AI Music Startup Suno Scrapes Millions of Songs
Suno, an AI music startup, has been caught using millions of songs, lyrics, and podcasts from platforms like YouTube Music and Deezer to train its models. This discovery followed a security breach by a hacker who exposed internal data and customer details. The dataset includes over 113,000 hours of YouTube Music audio, sparking debates about potential copyright violations and data privacy issues. Suno maintains that its data usage is protected under fair use, but this claim is likely to face scrutiny from the music industry. The situation underscores the complex relationship between AI development and intellectual property rights, as companies navigate the legal landscape.
WIRED AI
© WIRED AIMarket & Regulationbusiness
Google Revamps Gemini AI Usage and Pricing
Google has shifted its Gemini AI usage model from counting requests to measuring computing power, meaning users might face limits based on task complexity rather than quantity. This change aims to better reflect the resource costs for Google, but it introduces uncertainty for users about when they might reach their limits. Subscription plans now offer varying levels of access, with higher tiers allowing more complex and frequent AI interactions. This adjustment highlights Google's focus on managing resources efficiently, though it may complicate the user experience. Users must now consider the complexity of their tasks when planning AI usage. The new system could lead to more strategic use of AI resources by consumers.
WIRED AI
© WIRED AIMarket & Regulationbusiness
San Francisco Demands Removal of AI 'Nudify' Apps
San Francisco's city attorney has taken decisive action against Apple and Google, demanding they remove apps that generate AI-created nonconsensual nude images. These apps exploit AI technology to produce explicit deepfakes, often targeting women and girls, and have been downloaded millions of times. The legal notices accuse the tech giants of profiting from these harmful apps, which violate California laws against deepfake pornography. Despite existing policies, both companies have faced challenges in keeping such apps off their platforms, highlighting the need for more effective moderation. This move reflects growing concerns over the misuse of AI technology to create explicit content without consent, and pressures tech companies to take stronger action. The persistence of these tools calls for a reevaluation of current enforcement mechanisms to prevent future occurrences.
WIRED AI
More in Research
© Google DeepMindResearchresearch
Google DeepMind and Isomorphic Labs Tackle Bioresilience
Google DeepMind and Isomorphic Labs are taking a proactive stance on bioresilience by leveraging AI to enhance global biosecurity. Their approach involves preventing misuse of AI models while empowering governments and scientists to respond to biological threats. With tools like AlphaFold and IsoDDE, they aim to accelerate the discovery of therapeutics and improve pathogen detection. By making AI models available to trusted partners, they focus on prevention, detection, and response to outbreaks, aiming to safeguard global health with precision and speed.
Google DeepMind
© MIT News AIResearchresearch
MIT Develops System to Enhance CAD Model Generation
MIT researchers have developed a novel system that significantly improves the conversion of 2D designs into 3D CAD models using vision-language models. This system, called GIFT, enhances the accuracy and functionality of CAD programs while reducing computational demands. By learning from its own errors, the system generates new data to refine its performance, offering a more efficient and cost-effective approach to rapid prototyping. This advancement could transform how engineers approach design, making AI-driven CAD generation more reliable and accessible for everyday engineering tasks.
MIT News AI
© MIT News AIResearchresearch
MIT Introduces 'Neural Transparency' for AI Design
MIT researchers have introduced 'neural transparency,' a tool that allows users to visualize an AI's neural network behavior before interaction. This innovation aims to address the common issue where users misjudge their AI's behavior, often overestimating positive traits and underestimating negative ones. By providing a 'brain scan' of AI, users can anticipate potential risks during the design phase rather than after deployment. This approach could shift AI design from reactive to proactive, helping users create more reliable and transparent AI companions. However, while transparency increased trust, it didn't change design practices, indicating further work is needed to influence user behavior.
MIT News AI