Fireship thumbnail

📝 Fireship Blog

The Rise and Fall of T: What the Massive Data Breach Teaches Us About App Security

Last week, a groundbreaking dating app designed exclusively for women, known as T, rocketed to the top of the App Store charts—only to implode spectacularly following one of the most embarrassing data breaches in recent memory. T’s unique concept was simple yet powerful: women could anonymously share stories, photos, and warnings about men they’d dated, helping others avoid bad experiences. However, this noble intention was undone by catastrophic security failures that compromised the privacy of thousands of users.

What Was T and Why Did It Matter?

T targeted a very specific niche in the online dating world. By allowing women to "dox" and gossip about men they dated, the app aimed to serve as a protective shield against bad actors on popular dating platforms like Tinder. Given that a small percentage of highly desirable men attract the majority of attention on these apps, T’s database likely contained candid accounts about some of the most sought-after individuals.

To use the app, women had to verify their identity by submitting a selfie alongside a government-issued ID, ensuring that only verified women could participate. While this verification step was intended to maintain community safety, it became a double-edged sword when the data was exposed.

How Did the Breach Happen?

On July 25th, T confirmed that an unauthorized party had accessed a legacy Firebase storage bucket critical to their infrastructure. This bucket was left completely unsecured—unencrypted and publicly accessible on the internet—making it shockingly easy for hackers to retrieve data. Approximately 72,000 images—including 13,000 selfies and ID photos—were compromised.

But the damage didn’t stop there. A few days later, another database was leaked containing more than 1.1 million posts, comments, and private messages. The fallout was immediate and severe, with personal data of thousands of women exposed and spread across platforms like 4chan.

The Fallout: Privacy Violated and Community Mocked

The leaked data led to an ironic and painful twist: the very community T aimed to protect found itself victimized. Users’ selfies and sensitive information flooded the internet, and instead of sympathy, many faced ridicule and harassment.

Online communities quickly seized the opportunity to mock the users, derogatorily nicknaming them “roasties.” Developers and hackers alike began experimenting with the exposed data, creating everything from detailed data analyses in Python to interactive maps plotting user location data in JavaScript. Some even built websites ranking users by their appearance—a disturbing invasion of privacy fueled by this careless breach.

The App’s Response: A Hollow Apology

In the wake of the breach, T’s team issued a statement that many saw as a non-apology filled with corporate jargon. They admitted that a legacy storage system was penetrated but failed to fully take responsibility or explain how such a critical security lapse could have occurred.

What made the breach especially egregious was that Firebase—the storage service used—warns repeatedly when buckets are left public. There are built-in alerts and email reminders indicating when data is accessible to anyone online. Yet, T’s developers ignored or missed these warnings. Notably, the app’s user interface promised that selfies would be deleted post-verification—a promise that turned out to be false.

Lessons Learned: The High Cost of Incompetence in App Development

Interestingly, the app was developed by Shawn Cook, a male coder with less than six months of professional experience according to LinkedIn. While some speculate that the app was a rushed “vibecoded” project lacking polish, the real takeaway is a cautionary tale about security negligence.

Firebase’s infrastructure is known for its ease of use, but with that comes the responsibility to configure security rules correctly. Leaving such sensitive data exposed isn’t just a mistake—it’s a fundamental failure in understanding or prioritizing user privacy.

The Bigger Picture: Data Security in a Changing Digital Landscape

The T breach underscores broader concerns about data protection, especially as governments implement stricter age verification laws for adult content and online safety. Apps requiring ID verification must be especially vigilant, as their data stores contain some of the most sensitive personal information imaginable.

The reality is that even well-intentioned apps can become liabilities if security isn’t baked into their design from the start. For developers, this means investing time and resources into proper data encryption, access controls, and regular audits.

How Developers Can Avoid Being the Next T

If you’re a developer or startup founder, the T breach is a stark reminder of why infrastructure choices matter. Firebase and similar platforms provide powerful tools—but they demand careful configuration.

For those looking to maintain control and security over their projects, consider alternatives like virtual private servers (VPS). Services like Hostinger offer affordable, flexible VPS hosting with multiple operating systems and pre-installed templates for popular frameworks. This approach gives developers full ownership of their environment and reduces the risk of accidental exposure due to misconfigured cloud services.

Final Thoughts

T’s meteoric rise and catastrophic fall is a cautionary tale for the tech world. A revolutionary concept meant to empower women was undone by poor security practices, ultimately harming the very community it sought to protect.

As users, always be aware of the data you share online and the risks involved. As developers, never underestimate the importance of securing your data—because once it’s out, there’s no taking it back.


Stay informed with the latest in tech and security by following our blog. If you’re a developer seeking reliable hosting solutions, check out Hostinger for great performance and control over your projects.

Alibaba’s Quen 3 Coder: A New Challenger in AI-Powered Coding

In a major breakthrough for AI-driven programming, Chinese tech giant Alibaba has unveiled Quen 3 coder, an openweight, long-horizon mixture of experts agent coding model that is already making waves in the developer community. Released on July 23rd, 2025, Quen 3 coder is the first openweight model to rival the programming prowess of Claude 4, currently regarded as the leader in AI coding tools.

What Makes Quen 3 Coder Stand Out?

Massive Training with a Code-Heavy Diet

Quen 3 coder was trained on an astonishing 7.5 trillion tokens, with 70% of that data being code—a volume far beyond what even the most seasoned developers have seen in a lifetime. To put it into perspective, this is equivalent to a billion times more code than a developer with 50 years of experience.

Advanced Training Techniques

The model’s training process is highly sophisticated, leveraging long horizon reinforcement learning across 20,000 parallel environments. This means Quen 3 coder learns by actively solving real-world coding problems, executing, and testing code simultaneously—like a coding boot camp with thousands of tireless graduates all working in unison.

Impressive Performance Benchmarks

Benchmarks reveal that Quen surpasses notable models like Kimmy K2 and GPT-4.1, and approaches the elite performance level of Claude 4, despite having a significantly smaller model size. Smaller models are crucial because they require less computational power and electricity, making them more efficient for deployment.

Unmatched Context Window

One of Quen 3 coder’s most striking features is its massive 256,000 token context window, which can stretch up to 1 million tokens. This capacity can easily encompass entire codebases of startups, including all their technical debt, enabling it to understand and work with large, complex projects seamlessly.

Quen CLI: Command-Line Power for Developers

Alongside the model release, Alibaba introduced a new CLI tool, forked from the open-source Gemini CLI. This tool fully exploits Quen 3 coder’s agentic capabilities, allowing developers to run, execute, and test code directly from the command line. This integration marks a significant step toward making AI coding assistants more practical and accessible in everyday programming workflows.

Practical Considerations: Accessibility and Usage

Despite its groundbreaking capabilities, Quen 3 coder’s full 480 billion parameter version demands enormous resources—requiring tens or hundreds of thousands of dollars in GPU infrastructure and substantial electricity costs. Hence, running it locally on a typical laptop is unrealistic.

Most users will likely access Quen 3 coder via API keys from cloud providers and use the new Quen CLI to interact with the model. This cloud-based approach democratizes access while maintaining the heavy computational lifting on powerful servers.

The AI Coding Landscape: Quen vs. Claude and Others

While Quen 3 coder represents a significant leap for open coding models, it remains to be seen if it can dethrone Claude 4’s dominance. To truly topple Claude, a new model must be not only open and affordable but also deliver a decisive leap in coding capabilities.

OpenAI, despite recent setbacks like talent losses and delays in open model releases (possibly due to competition from Chinese models like Quen), continues to make strides. Notably, OpenAI and Google both recently achieved gold medals in the International Mathematical Olympiad, showcasing their AI’s prowess in complex problem-solving—another domain relevant to advanced coding skills.

Boost Your Coding Today with Code Rabbit

For developers eager to harness AI for better coding now, Code Rabbit offers a powerful solution. Their free VS Code extension provides advanced code reviews directly in your editor, plus a new “fix all with AI” feature that applies all review suggestions automatically through your AI coding agent of choice.

Code Rabbit saves you from manually addressing each review comment, freeing up time to focus on writing more code—flaws and all! It integrates seamlessly with VS Code and popular forks like Cursor and Windsurf. You can download it for free using the link below.


Final Thoughts

Alibaba’s Quen 3 coder pushes the boundaries of open AI models in coding, combining enormous training data, cutting-edge reinforcement learning, and a massive context window to challenge industry leaders. While it may not unseat Claude 4 just yet, it signals a new era of powerful, accessible AI coding assistants.

As AI continues to evolve, developers have more tools than ever to enhance their productivity and creativity. Whether through giants like Quen 3 coder or practical extensions like Code Rabbit, the future of programming looks increasingly collaborative between humans and intelligent machines.


Try Code Rabbit today and supercharge your coding workflow!

This has been The Code Report. Thanks for reading, and happy coding!

Amazon Launches Kira: A New AI-Powered IDE Challenging the Developer Tool Landscape

Yesterday, Amazon surprised the developer community by releasing its own AI-powered Integrated Development Environment (IDE) named Kira. This new tool enters a crowded field of AI-assisted coding editors, competing alongside notable names like Cursor, Windsor, Firebase Studio, and GitHub Copilot. What sets Kira apart is its foundation on Claude Sonnet 4.0 and a unique approach called spec-driven development, which promises better handling of complex projects compared to its rivals. And for now, it’s completely free to use.

The Growing AI IDE Ecosystem

The AI coding assistant market is rapidly evolving. Just recently, Chinese developers introduced Kimmy K2, a lightweight, agentic coding model that rivals Claude in performance. These advancements hint at a future where AI tools might even automatically fix bugs in pirated software—potentially ending the endless internet ridicule of poorly maintained codebases.

In the backdrop of this surge, Amazon’s Kira aims to become a serious contender. The timing is notable: earlier this year, OpenAI’s planned $2 billion acquisition of Windsor, another popular VS Code fork, collapsed. Google then swooped in, recruiting Windsor’s key talent for $2.4 billion, leaving many employees in limbo. Meanwhile, Cognition, the company behind another IDE called Devon, acquired the remaining Windsor assets.

This frenzy underscores a simple truth: companies are willing to pay billions for developer tools that can capture the hearts and workflows of programmers. And the biggest winner so far has been Anthropic, mainly due to the success of its Claude Cloud Code CLI tool. Anthropic’s revenue skyrocketed from under $1 billion to over $4 billion in just one year, with Amazon as a major investor, having poured $8 billion into the company.

The Cursor Dilemma and Kira’s Opportunity

If you’re a user of Claude-powered IDEs, chances are you’ve tried Cursor. While Cursor is powerful, it’s heavily dependent on Anthropic’s Claude model, which puts the startup in a tight spot: it must price its services to eventually turn a profit while footing the bill for Claude’s usage. This led to recent controversial pricing changes that caught many users off guard, resulting in public apologies from Cursor.

Enter Kira, Amazon’s VS Code fork with similar features but a more attractive pricing model that could offer more value for less money. Whether this is a coincidence or a strategic move to undercut Cursor remains unclear, but it certainly shakes up the market.

Hands-On With Kira: A Different Approach to AI Coding

After testing Kira, the user interface impresses with its clean design but feels somewhat slower than competitors. Some expected early-stage bugs appeared, such as missing features like chat checkpoints and occasional delays due to server overload.

The most distinctive feature of Kira is its spec-driven development workflow. Instead of rushing to generate code immediately, Kira encourages developers to begin with a requirements markdown file that outlines user stories and acceptance criteria. This is followed by a design document detailing implementation plans, including component structure, testing strategies, and error handling.

Only after these planning stages does Kira proceed to actual code generation, combining requirements and design into a comprehensive implementation plan. This step-by-step process may feel slower but is ideal for serious software projects or team environments where design documentation and consistent code quality are critical.

Who Is Kira For?

Kira seems tailored for enterprise developers who value thorough planning and collaboration over rapid prototyping. Currently, the IDE is closed-source and exclusively powered by Claude, but Amazon plans to integrate other AI models in the future.

Since Kira is free to use at launch, it’s worth trying out—especially if you’re looking for an AI coding assistant that emphasizes structured development and quality assurance. Whether it can dethrone established players like Cursor remains to be seen.

Learn the Foundations Behind AI Coding Tools

To truly leverage the power of AI in software development, understanding the underlying math and computer science is essential. For those eager to dive deeper, Brilliant.org offers an excellent hands-on course called “How AI Works”, teaching you how to build a language model from scratch and experiment with advanced concepts like feature vectors.

Brilliant’s interactive approach is proven to be six times more effective than passive video lectures. You can try all their courses free for 30 days by visiting brilliant.org/fireship or scanning the QR code below. Plus, you’ll get 20% off an annual premium subscription.


In summary, Amazon’s entry into the AI IDE space with Kira marks an exciting new chapter in developer tooling. With its spec-driven methodology and strong backing, Kira could reshape how coding AI assists teams in building complex software. As always, the race to win developers' loyalty continues—and that means better tools and experiences for everyone.

Thanks for reading The Code Report. Stay tuned for more updates on the future of coding and AI.


Disclosure: This post is sponsored by Brilliant.org.