The Real Software Security Problem Is Us
A colleague recently described a fun Friday night for his teenage son: staying home and chatting online. Every now and then there’s a party where all his friends talk on their laptops.
We’re getting more and more used to living in such virtual spaces—yet there’s an important distinction to be made between virtual spaces and the actual physical spaces we walk around in. We usually expect that we can walk across a bridge or into a building without the structure collapsing. We don’t have that kind of confidence with software programs.
Believe it not, we could, in the not-so-distant future, actually live in a world where software doesn’t randomly and catastrophically fail. Our software systems could withstand attacks. Our private social media and health data could be seen only by those with permission to see it. All we need are the right fixes.
The problem with modern software is that we’ve been building our “skyscrapers” with the same materials and techniques used to build huts. Software began as a collection of bricks: simple procedures, sequences of commands for calculations, games, and curiosities. Decades later, we have millions of procedures interacting with each other on interconnected machines with access to all kinds of secret information. Yet we’re still using similar languages and tools.
If we want our systems to have fewer vulnerabilities, we need to use better building materials. The languages people use today make it too easy for the programmer to make mistakes, and they make it too hard to detect the mistakes.
A better way would be to use languages that provide the guarantees we need. The Heartbleed vulnerability happened because someone forgot to check that a chunk of memory ended where it was supposed to. This could only happen in a programming language where the programmer is responsible for managing memory. So why not use languages that manage memory automatically? Why not make the programming languages do the heavy lifting?
Another way would be to make software easier to analyze. Facebook had so much trouble making sense of the software it used that it created Hack and Flow, annotated versions of PHP and Javascript, to make the two languages more comprehensible.
This is partly our own fault. We find online life fun, so we tend to let sites do whatever they want with our personal data. Software companies respond by churning out new features as quickly as possible, using the most convenient materials and tools at the expense of security.
Change won’t happen until we demand that it happens. Our software could be as well-constructed and reliable as our buildings. To make that happen, we all need to value technical soundness over novelty. It’s up to us to make online life is as safe as it is enjoyable.
Jean Yang is an assistant professor of computer science at Carnegie Mellon University, and the cofounder of Cybersecurity Factory, an accelerator focused on software security.
Keep Reading
Most Popular
Large language models can do jaw-dropping things. But nobody knows exactly why.
And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.
The problem with plug-in hybrids? Their drivers.
Plug-in hybrids are often sold as a transition to EVs, but new data from Europe shows we’re still underestimating the emissions they produce.
Google DeepMind’s new generative model makes Super Mario–like games from scratch
Genie learns how to control games by watching hours and hours of video. It could help train next-gen robots too.
How scientists traced a mysterious covid case back to six toilets
When wastewater surveillance turns into a hunt for a single infected individual, the ethics get tricky.
Stay connected
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.