Research
Concerns Rise Over OpenAI's Astra Model Release
The Verge AIhigh confidence
Why it matters
- →Astra's opaque architecture could hinder the ability to monitor AI behavior effectively.
- →The release raises concerns about a competitive 'race to the bottom' in AI safety standards.
- →OpenAI's approach highlights the ongoing tension between AI innovation and safety oversight.
OpenAI is preparing to release its Astra model, which has raised alarms among AI safety researchers due to its opaque architecture. Astra employs a recurrent depth technique, making its internal reasoning less visible compared to traditional models. This has led to concerns about the difficulty of monitoring the model's behavior and the potential for a 'race to the bottom' in AI safety standards. OpenAI has stated it will use additional monitoring to address these issues, but the situation underscores the challenges of balancing AI advancement with safety oversight.
Read originalMore from The Verge AI
© The Verge AIModels & Labsmodels
Google Launches Gemini 3.8 Flash Model
Google's new Gemini 3.8 Flash model promises enhanced performance by executing more reasoning steps and iteratively calling tools, potentially increasing token usage and costs. Despite maintaining the same introductory pricing as its predecessor, the model's efficiency in handling complex tasks could lead to higher expenses for users. It has shown significant improvements in software engineering and autonomous AI agents, outperforming competitors on various benchmarks. The model is available to consumers and developers, with a special version, Gemini 3.8 Flash Cyber, offered to governments and trusted partners through Google's Fairwind Program.
The Verge AI
© The Verge AIGeneral AIwriting
Amazon's Alexa Can Now Spot Fake Emails
Amazon has enhanced its AI assistant, Alexa, to help users identify fake emails, texts, and calls claiming to be from the company. This new feature allows users to ask Alexa for Shopping to verify messages by comparing them against Amazon's official records. If Alexa confirms a message is genuine, it assures the user; otherwise, it advises checking orders in the app or contacting support. This development aims to combat impersonation scams, providing users with a more secure shopping experience. It's a significant step in using AI to enhance consumer protection against fraud.
The Verge AI
© The Verge AIMarket & Regulationbusiness
Trump Administration Backs OpenAI in NYT Copyright Case
In a significant legal development, the Trump administration has thrown its support behind OpenAI in a copyright lawsuit filed by The New York Times. The administration argues that training AI models on copyrighted material should be considered fair use, emphasizing its importance for American prosperity and scientific progress. This intervention could influence the outcome of the case, which seeks billions in damages from OpenAI and Microsoft. The decision could set a precedent for how AI labs interact with copyrighted content, impacting future media and AI industry relations.
The Verge AI
More in Research
© TechCrunch AIResearchresearch
OpenAI's Astra Model Raises AI Safety Concerns
OpenAI's Astra model introduces a new reasoning technique known as 'recurrent depth,' which has sparked significant concern among AI safety experts. This approach, also referred to as 'opaque recurrence,' allows the model to process queries in a loop, making its reasoning process less transparent and more challenging to monitor. Despite OpenAI's assurances that Astra's use of this technique is limited and that they remain committed to chain-of-thought monitoring, experts worry about the potential for diminished transparency in AI reasoning. The situation underscores the ongoing tension between advancing AI capabilities and ensuring safety and accountability in AI systems.
TechCrunch AI
© WIRED AIResearchresearch
Russian Startup Enables AI Models to Communicate Silently
Mostik, a Russian startup, has developed a novel approach allowing AI models to communicate without generating text output, akin to machine telepathy. This technique leverages the mathematical values in model weights to enable smaller models to benefit from the capabilities of larger ones, enhancing efficiency and performance. By creating a bridge between models like GLM-5.2 and Qwen-3.5, Mostik has demonstrated a cost-effective hybrid system that performs impressively. This innovation could significantly boost the value of open-weight models, challenging the dominance of proprietary models from major labs like OpenAI.
WIRED AI
Researchresearch
Motional and MIT enhance self-driving car transparency
Motional and MIT have developed a system that allows self-driving cars to explain their decisions in real-time, addressing the black-box problem in autonomous vehicle AI. Their Concept-Wrapper Network (CW-Net) translates the neural network's internal logic into human-readable concepts, providing transparency into the vehicle's decision-making process. This innovation was tested on public roads in Las Vegas, revealing insights into the car's behavior that were previously hidden. By making AI decisions more interpretable, this system could become a standard requirement as autonomous technology expands into new markets.
AI News