Security and privacy can sometimes be at odds, or they can go hand in hand, but finding enough of both requires nothing less than a fight. Whether you’re running security at a bank or protecting yourself from phone scammers and data-hungry apps, the bad guys have never been more plentiful, more capable, and more ubiquitous. With the abundance of digital devices and the help of advances in generative AI, the attack surface feels like everything, everywhere, all at once. But thanks in part to a booming cybersecurity industry (and no small amount of hype), the cyberdefenses are growing too, getting more sophisticated and easier to use.
AnjunaFor helping to improve confidential computingGenerative AI, which has raised expensive legal and ethical questions about who owns what data for what purpose, also looks like a problem for confidential computing, which aims to protect apps and data at all times, even while they’re in use, by keeping them inside a secure enclave that uses hardware-based methods to control access. One problem: All the varying and competing approaches by different companies aren’t always compatible. Founded by veterans of Israel’s Unit 8200, Anjuna attempts to tear through the mess by virtualizing confidential computing CPUs within a single software stack, thus avoiding the need to recode individual applications for every server infrastructure and cloud service. The approach has helped the company raise $42 million to date and find customers in banking, defense, and blockchain.
OpteryFor scaling data privacy demandsOptery, launched in California in 2020, uses a patented search tool to pierce data brokers’ anti-bot systems and find out who has your information, with not surprising but still jaw-dropping results. For a fee, its automated system will file requests on your behalf and perpetually keep tabs on who’s keeping tabs on you. The company has raised $6 million so far and envisions evolving into a personal data management platform that would instead allow customers to license and sell their data if they wanted to.
PindropFor keeping our voices to ourselvesWhile it’s getting harder for us to tell when a voice or a video is a fake, it can be a snap for AI, especially when you use the arsenal that Pindrop has: a combination of peer-reviewed and patented methods for detecting liveness in synthetically generated speech, recorded-voice replay attacks, and morphed voices, in addition to audiovisual deepfakes. Pindrop also relies on graph analysis, using more than 1,000 contextual data points, for instance, to identify at-risk accounts up to 60 days in advance of attempted account takeover or to simply verify that the person speaking is who they say they are. Think of it like a passive Captcha for voice—if Captchas could actually still stop bots.
The companies behind these technologies are among the honorees in Fast Company’s Next Big Things in Tech awards for 2023. See a full list of all the winners across all categories and read more about the methodology behind the selection process.
Accedi per aggiungere un commento
Altri post in questo gruppo
A software engineer became X’s main character last week after being outed as a serial moonlighter at multiple Silicon Valley startups.
“PSA: there’s a guy named Soham Parekh (in India) w
The flash floods that have devastated Texas are already a difficult crisis to manage. More than 100 people are confirmed dead
Amazon is extending its annual Prime Day sales and offering new membership perks to Ge
How would you spend $342 billion?
A number of games called “Spend Elon Musk’s Money” have been popping up online, inviting users to imagine how they’d blow through the
On Tuesday, AI lab Moonvalley
As Elon Musk announced plans over the Fourth of July weekend to establish a third political party,
