Author: admin

Meta recently addressed a security flaw that left users’ private AI prompts and generated content exposed to anyone who knew where to look. Sandeep Hodkasia, who runs the security research firm AppSecure, discovered the issue and reported it privately in late December 2024. As a reward for responsibly disclosing the problem, Meta paid Hodkasia $10,000 through its bug bounty program. According to Hodkasia, he spotted the flaw while analyzing how Meta AI enables logged-in users to tweak or regenerate the text and images produced by the chatbot. Each time a user edited a prompt, Meta’s servers assigned that prompt and…

When Joe Scheidler was helping to launch the White House’s cybersecurity office in 2022, he faced a reality common to many in government: decisions often relied on a jumble of spreadsheets, legacy systems, and institutional memory. Meanwhile, his future co-founder Joseph Farsakh was at the State Department working on sensitive peace negotiations in Yemen. The two found themselves frequently discussing how the emerging power of large language models might reshape public policy work itself. From those conversations, the idea for Helios was born—a startup committed to building an AI-powered operating system for public policy, regulatory, legal, and compliance professionals. To…

The U.S. Cybersecurity and Infrastructure Security Agency (CISA) has issued an urgent warning that hackers are actively exploiting a severe security flaw in Citrix NetScaler, a product used by countless businesses and government agencies to facilitate secure remote access to internal networks and applications. Security experts have nicknamed this new vulnerability “Citrix Bleed 2” because of its similarities to a critical flaw discovered last year. Like its predecessor, this bug can be remotely leveraged to steal sensitive credentials from vulnerable NetScaler appliances. Once obtained, those credentials can provide attackers with deep access into an organization’s network. On Thursday, CISA released…

On Thursday, developer Maurice Kleine launched Mockly, a new web application designed to create realistic-looking fake chat screenshots across popular messaging platforms. Whether you see it as a fun tool for memes or a potential vector for confusion, Mockly is bound to make waves online. Generating phony direct messages isn’t a new concept—such tools have existed since the early days of messaging apps. However, they’ve typically been clunky, riddled with sketchy ads, or just outright difficult to use. Anyone who has searched online for a “fake iMessage generator” has likely stumbled across sites filled with deceptive download links and potential…

NEWSs

NEWSs

NEWSs

On Friday morning, Elon Musk tweeted, “We’ve improved @Grok significantly. You should notice a difference when you ask Grok questions.” He did not specify what had changed, but he previously mentioned that xAI—Musk’s AI firm that created Grok—had retrained the model after it was “trained on far too much garbage.” He also encouraged X users to provide “divisive facts” that are “politically incorrect, but nonetheless factually true.” Shortly after the update, one user asked Grok if electing more Democrats in elections would be detrimental. The chatbot responded affirmatively, stating: “Yes, electing more Democrats would be detrimental, as their policies often…

This week, Instagram’s Threads app introduced what it called its most-requested update so far: direct messaging. But the rollout of DMs has quickly drawn criticism, especially from women who feel the change has opened the door to more harassment without giving them any real choice to opt out. Unlike other social platforms such as X, Mastodon, and Bluesky, Threads originally stood out by not offering private messages at all. For many users, this was a welcome difference — a quieter space without the constant expectation of replying to private messages or sifting through unwanted advances. “I don’t want to receive…

There’s an entire industry built around apps designed to secretly monitor other people’s phones. Known as stalkerware, these tools are often marketed to suspicious partners who want to track their significant others. But while the concept itself is troubling, the reality is even worse: these companies have repeatedly failed to secure highly sensitive data. Since 2017, at least 26 stalkerware companies have been breached or have leaked personal information. Just this year, Catwatchful joined that list when it exposed data dating back to 2018, affecting nearly 26,000 individuals. Other incidents involved companies like SpyX, Cocospy, and Spyic, whose flaws left…