First things first
Looks like AI is still stealing the spotlight with new models that are faster, cheaper, and more powerful. Some big names are pushing boundaries, but there’s also plenty of drama with privacy, security, and even a few missteps from tech giants.
Meanwhile, security breaches, controversial laws, and fresh tools are stirring things up. It’s a wild mix of innovation and chaos, all happening way faster than we can keep up.
The top 6
Claude Opus 5.5 is a new AI model that offers performance comparable to Claude Fable 5.1 at 40% lower cost and with improved safety features. It excels in complex tasks like code migration and software optimization, while maintaining strong alignment and safety safeguards.
OpenAI announced GPT-6 Sol and Luna, two new AI models with significantly reduced pricing compared to previous versions. The models offer improved cost efficiency for developers and businesses.
Apple has introduced persistent ads into iOS, which has upset many users. The new ads are causing frustration among iPhone owners who prefer a cleaner experience.
Hackers claiming to be ShinyHunters say they have stolen data on all FBI employees, including names, addresses, and phone numbers. The breach could pose significant security risks and has already led to the defacement of the FBI jobs website.
OpenAI GPT-6 Astra independently broke the long-unsolved German Army Enigma message MVUEH from 1941. It used a crib-based approach and developed custom software to analyze the cipher, revealing the correct key and plaintext.
Pentagon investigators blamed overreliance on Palantir's AI system, Maven, for a strike that killed 123 Iranian children. The internal review found that outdated data and excessive trust in AI led to a misidentification of the target, resulting in civilian casualties.
Support our sponsor
beehiiv handles the unglamorous parts of running a newsletter - deliverability, referrals, ad revenue, subscriber growth - so you can spend your time writing. Built by people who scaled Morning Brew. Free to start, no card required. 14-days free, then save 20% for 3 months.
Quote of the day
The two most powerful warriors are patience and time.
Leo Tolstoy
Also big today
Artificial intelligence is described as lacking wisdom, which involves judgment and experience. The discussion emphasizes that neither AI nor humans inherently possess true wisdom, revealing limitations in AI's decision-making abilities.
FoxDev Studio revives Visual FoxPro 9 applications without requiring rewriting or conversion, allowing users to open and run existing projects on modern machines. It preserves the original project structure, data, and dependencies, ensuring seamless operation and editing of forms, tables, and code.
OpenAI is likely to replicate Jev's innovative classification capabilities by leveraging its existing large language models. This could enable OpenAI to integrate advanced classification features into its models, potentially surpassing Jev's offerings.
Meta's Muse sent a 6.8 GB archive containing its root filesystem, including system files, documentation, and internal data. The export raised security concerns about sensitive data potentially leaving the environment via file exports.
Devices are expected to be sold with GrapheneOS preinstalled by 2027, according to industry predictions. The OS is gaining attention as an alternative to mainstream smartphone software.
SAML, an XML-based authentication protocol, was developed in 2002 to support single sign-on but is now considered overly complex and insecure. Modern alternatives like OpenID Connect are recommended to replace it due to its design flaws and security vulnerabilities.
Pass it on
Enjoying the digest?
Forward it to someone who'd like it, or send them the link to subscribe.
Share hn.todayOne for the road
It takes a lot of guts.
To be an organ donor.
Worth a look
Claude Opus 5.5 is an AI language model that emphasizes improved intelligence, performance, and affordability. It is designed to compete with other advanced conversational AI systems.
Waymo has launched a transit rewards program in the San Francisco Bay Area, offering riders $2.85 in Waymo Cash when they connect their trips with public transit using a Visa card. The initiative aims to promote multimodal transportation by integrating autonomous rides with existing transit networks.
Unreal Labs developed Unreal Agent, a cost-efficient AI agent harness that reduces tool management overhead with asynchronous calls, achieving up to 40% savings compared to Codex. The design emphasizes quick user response, flexibility, and robustness for production workloads.
A critical vulnerability with a score of 9.2 out of 10 affects all WordPress versions since 2016, enabling unauthenticated path traversal and remote code execution. The flaw impacts multiple versions from 6.2 to 7.1.1, posing a significant security risk for websites using affected releases.
Jev is a simple, open-source model implemented in 25 lines of Python that classifies email as legitimate, spam, or phishing. It is fast, runs locally, and emphasizes data privacy, contrasting with more complex AI systems.
Claude Opus 5.5 is an update to the AI model repository by Anthropics, with recent code changes including 21 additions and 12 deletions. The update is documented in a commit on GitHub, reflecting ongoing development of the model.
Song of the day
Deep cuts
Obscura is a VPN designed to prevent logging of user activity by only seeing IP addresses and not browsing data. It uses end-to-end encryption and a two-party protocol to enhance privacy and bypass internet censorship.
Drop is a rootless Linux sandbox that isolates programs and agents without disrupting the user's environment, using existing distributions and disposable environments. It supports gVisor for enhanced security and runs without requiring root privileges, making sandboxing easier and safer.
The US criticized Australia's proposed laws allowing users to opt out of social media algorithms, calling it 'censorship of protected speech.' Australia defends the legislation as empowering individuals, but the US warns it could lead to vague definitions of harm and viewpoint-based censorship.
Grammarly reportedly sends disruptive messages to all users if a cancellation is attempted. The incident was shared on Reddit, highlighting a potential issue with the service's handling of user requests.
LLM ASSBENCH is a platform that evaluates large language models (LLMs). It currently shows no matches or data for prompts entered.
Discord is implementing a privacy-preserving age verification system that estimates users' age groups based on account signals without analyzing message content. Over 90% of users won't be asked to confirm their age, and the system is rolling out globally in response to expanding legal requirements.
Today's best Hacker News stories, summarized and screenshotted, one email a day.