ContentBuffer
OpenAI's Astra model has passed a critical cybersecurity test, scoring a perfect score on ExploitBench and discovering two zero-day vulnerabilities. This model can find and exploit security flaws without human guidance, raising concerns about its potential misuse. OpenAI is taking precautions, including chain-of-thought monitoring and restricting responses to higher-risk accounts. The company plans to release more evaluations and safety information before a wider launch. Developers and security teams need to stay alert as this could change how we approach AI security.
Google just dropped five new updates to Android 17, aimed at making phones more accessible and personalized. The star of the show is Motion Assist, a feature that reduces motion sickness by adding a screen overlay in moving vehicles. It's a big deal for anyone who gets queasy while navigating with their phone. Guided Vision, another highlight, helps blind and low-vision users with tasks like reading small print or identifying objects. It's rolling out to Android 9 and newer in Gemini-supported countries. Plus, Find Hub now remembers important items, and Google Messages gets a Google Keep integration and customizable chat backgrounds. These updates are a direct response to Apple's similar features, leveling the playing field for Android users. Let's dive deeper…
In today's ContentBuffer update:
X Money Users Reporting Password Reset Email Flood
Apple CEO Transition and Upcoming iPhone Event
Anthropic launches Fable 5.1 and Mythos 5.1, setting new benchmarks
Thrive Capital Withdraws from FIFA's $20B Corporate Deal
5 new AI tools & 5 new AI jobs
More tech news
Latest Development
Image source: techcrunch.com
Summary: OpenAI's Astra model passed a critical cybersecurity test, finding and exploiting zero-day vulnerabilities. The company is taking safety measures to prevent misuse. Developers and security teams should watch closely.
Key Points:
Astra scored a perfect score on ExploitBench, a cybersecurity test.
The model discovered and exploited two zero-day vulnerabilities in the test.
OpenAI is implementing chain-of-thought monitoring to detect and stop bad behavior.
The company is restricting model responses to higher-risk accounts.
OpenAI expects to release more evaluations and safety information before a wider launch.
Why it matters: If you're working on cybersecurity or AI safety, Astra's performance on ExploitBench and its ability to find zero-day vulnerabilities without human guidance are a game-changer. This could force a rethink of how we secure systems against AI-driven threats. The model's potential for misuse also means security teams need to adapt their strategies to prevent jailbreaks and abuse.
Image source: techcrunch.com
Summary: Google adds 5 new accessibility features to Android 17, including Motion Assist for motion sickness and Guided Vision for blind and low-vision users. Why it matters: Android users now have more options to match iPhone's accessibility.
Key Points:
Motion Assist reduces motion sickness with a customizable screen overlay for Android 17 users.
Guided Vision helps blind and low-vision users with tasks, available in Android 9 and newer in Gemini-supported countries.
Find Hub's 'remembered items' section lets users store and identify important objects with attached photos.
Google Messages gains Google Keep integration for sharing and editing information within conversations.
Customizable chat backgrounds, photos, and colors are coming to Google Messages for a more personal experience.
Why it matters: If you're an Android user with motion sickness, Motion Assist is a game-changer, reducing queasiness in moving vehicles. For blind and low-vision users, Guided Vision offers voice guidance for tasks like reading small print. Find Hub's 'remembered items' feature helps track and identify important objects, making daily life easier. These updates bring Android closer to iPhone's accessibility features, ensuring more users can enjoy the benefits of modern smartphones.
Image source: techcrunch.com
Summary: X Money users are getting hit with password reset spam. Enable two-factor auth ASAP.
Key Points:
X Money users report receiving multiple password reset emails since the service's launch.
Grok, X's chatbot, guides users to enable two-factor authentication for account security.
The company's general counsel has issued a warning against attackers attempting to victimize users.
No confirmed system breach or mass takeovers have been reported as of now.
X Money includes a bank card and other benefits, facilitating the company's digital economy.
Why it matters: If you're using X Money, you're likely receiving password reset emails. Enable two-factor auth to protect your account. The company's digital economy efforts could be hindered if users lose trust due to security concerns.
Image source: techcrunch.com
Summary: Apple's new CEO takes over, hosting a September 9 event with rumored foldable iPhone and AI advancements for Siri. Developers and tech enthusiasts should pay attention to this pivotal moment for Apple.
Key Points:
Apple's new CEO officially takes the helm, succeeding a 15-year leader.
iPhone event on September 9, rumored to include a foldable iPhone.
Siri AI improvements aim to compete with chatbots like Anthropic's Claude.
New products in the works include camera-equipped AirPods and smart glasses.
Previous CEO hands over with confidence in the new leader's capabilities.
Why it matters: Developers and product managers at Apple should closely follow the September 9 event for updates on Siri's capabilities and potential new APIs. The foldable iPhone could also impact app development, requiring teams to adapt their UI designs for a new form factor. Additionally, the introduction of camera-equipped AirPods and smart glasses could open up new opportunities for developers in the wearables and augmented reality space.
Image source: techcrunch.com
Summary: Anthropic released Fable 5.1 and Mythos 5.1, setting new benchmarks and offering exclusive access to partners. If you're in cybersecurity or life sciences, this could change your data handling game.
Key Points:
Fable 5.1 reduces token costs and enhances safeguards, available now on cloud platforms and through Anthropic API.
Mythos 5.1 is exclusive to registered Anthropic partners in cybersecurity and life sciences, rolling out now.
Both models set new records in Terminal-Bench 4.0 and Humanity's Last Exam, outperforming previous versions.
High-privacy services called Enterprise Frontier Safeguards will roll out to users in June, offering more control over monitoring.
Anthropic assures customers that their data has not been inappropriately accessed, addressing privacy concerns.
Why it matters: If you're in cybersecurity or life sciences, Mythos 5.1's exclusive access could change your data handling and compliance workflows. Fable 5.1's reduced token costs make it more accessible for general use, improving efficiency and cost-effectiveness for teams working with large datasets.
Image source: techcrunch.com
Summary: Thrive Capital backed out of a $20B deal with FIFA to privatize commercial aspects of FIFA competitions after backlash from European and North American football associations. The deal is now facing legal challenges and public outcry.
Key Points:
Thrive Capital agreed to be a lead investor in a potential sale of $4.2 billion worth of secondary shares in FIFA Forward Enterprise (FFE), representing 20% of the company.
The proposed company, FIFA Forward Enterprise (FFE), would have been fully owned by FIFA's member associations (MAs), with the option for MAs to sell shares to private investors.
UEFA, representing 55 of FIFA's 211 member associations, is pursuing legal action against FIFA president Gianni Infantino over the scrapped deal.
North American football associations, including Concacaf, also objected to the deal, joining European teams in the backlash.
Thrive Capital has engaged a lawyer, Alex Spiro, to represent it in the legal action, with documents on the deal's valuation being contested by UEFA.
Why it matters: UEFA and North American football associations objected to the privatization of FIFA competitions, fearing outsider influence. The legal challenges and public outcry have led to the cancellation of a $20B deal, highlighting the power of member associations in FIFA governance. The deal's valuation is now being contested, potentially affecting the financial interests of involved parties.
New Tools & Job
MachineTranslation.com - MachineTranslation.com is an AI-powered translation platform that translates text and documents by comparing outputs from multiple leading AI translation models. Its consensus technology helps users choose the most accurate translation, reducing errors and improving quality across more than 330 languages. What it does Translates text and documents using AI. Compares translations from multiple AI models. Provides a consensus translation to improve accuracy. Supports 330+ languages. Helps users evaluate and refine translations. What makes it good Higher accuracy: Compares multiple AI translations instead of relying on a single model. Better quality: Consensus technology helps identify the strongest translation. Fast and easy: Delivers results in seconds with a simple interface. Broad language support: Covers over 330 languages. Useful for professionals and travelers: Suitable for business, marketing, legal, technical, and everyday communication.
Format Factory - Format Factory is a browser-based workbench for common audio and video conversion jobs. It supports batch processing for video and audio conversion, compression, merging, audio extraction, and audio removal, with transparent credits and temporary results. Try Format Factory for practical media conversion without installer bundles or per-file setup.
HappyHorse.AI - HappyHorse.AI is a multimodal AI video generator for creators, marketers, and social media teams. Users can create videos from text prompts, animate images, or generate videos from reference characters. The platform focuses on browser-based video creation and native audio support.
NeatScribe - NeatScribe is designed for anyone who needs to reuse information from recorded speech. It creates timestamped transcripts, supports translation, and exports results as text files, documents, subtitles, captions, or timed lyrics.
Sakana Marlin - Sakana Marlin is Sakana AI's first commercial product: an autonomous 'Ultra Deep Research' agent positioned as a Virtual Chief Strategy Officer. Given a research topic, Marlin works autonomously for up to roughly eight hours, forming hypotheses, gathering and reconciling information from online sources, and synthesizing a detailed strategy report up to ~100 pages plus executive-summary slides. It builds on Sakana AI research including AB-MCTS (NeurIPS 2025 Spotlight) and The AI Scientist (published in Nature). A closed beta ran from April 2026 with ~300 professionals across finance and consulting.
Forward Deployed Engineer - Software Engineer - Spain - About ElevenLabs ElevenLabs is an AI research and product company transforming how we interact with technology. We launched in January 2023 with the first human-like AI voice model. Today, we serve millions of users and thousands of businesses - from fast-growing startups to large enterprises like D
Safety Engineer - About ElevenLabs ElevenLabs is an AI research and product company transforming how we interact with technology. We launched in January 2023 with the first human-like AI voice model. Today, we serve millions of users and thousands of businesses - from fast-growing startups to large enterprises like D
Engineering - Internal AI Transformation - About ElevenLabs ElevenLabs is an AI research and product company transforming how we interact with technology. We launched in January 2023 with the first human-like AI voice model. Today, we serve millions of users and thousands of businesses - from fast-growing startups to large enterprises like D
Research Engineer - About ElevenLabs ElevenLabs is an AI research and product company transforming how we interact with technology. We launched in January 2023 with the first human-like AI voice model. Today, we serve millions of users and thousands of businesses - from fast-growing startups to large enterprises like D
QUICK HITS
John Ternus Takes Over as Apple's CEO - John Ternus becomes Apple's CEO, taking over from Tim Cook. Ternus' long tenure and deep experience make him well-suited for the role.
Google Pics: New Image Creation Tool in Google Workspace - Google launches Pics, an AI-driven image creation tool for Google Workspace, aimed at designers and marketers. It's rolling out to Workspace customers and Google AI Pro/Ultra subscribers.
OpenAI Integrates ChatGPT Health with Epic's EHR for 325M Patients - OpenAI integrated ChatGPT Health with Epic's EHR, giving clinicians AI-powered insights from patient records. This could transform how healthcare providers use AI in patient care.
Waymo Debunks Pure AI for Autonomous Vehicles, Tesla Cybercab Nears - Waymo argues against pure AI for autonomous vehicles, pushing for sensor integration. Tesla's Cybercab, a purpose-built autonomous vehicle, is set to debut on September 3.
Empirik Raises $21M to Predict and Prevent Tech Outages - Empirik raises $21M to predict and prevent tech outages, helping DevOps teams offload routine troubleshooting.
GoPro Acquired by Starman Optical for $285M - Starman Optical acquires GoPro for $285M, wiping out $92M debt and giving GoPro shareholders $1.14 per share. The deal closes by year-end, with GoPro retaining 10% of the combined company.
THAT'S A WRAP
Our newsletter reaches a diverse audience of tech professionals, investors, engineers, managers, and business leaders worldwide. Get in touch today.
Be the first to comment
Get the daily edition delivered free every morning.
Join 3,451 builders reading daily.
Google introduces Motion Assist, Guided Vision, and more to Android 17, helping users with motion sickness, vision impairments, and item tracking. These updates bring Android closer to iPhone's accessibility features.
Orbify.eu is developing a curved map UI that combines street-level and overhead views, potentially improving navigation. The startup has a patent pending and is testing the technology with an interactive demo.
The Pentagon has launched custom versions of ChatGPT and Grok for government use, exempt from data collection and available to 1.7 million users. GenAI.mil offers secure access to commercial AI models without compromising sensitive data.