Author Archives: Poseidon

LaunchDarkly adds real-time controls for AI agents in production

LaunchDarkly has launched AgentControl, a new solution that gives software teams real-time control over AI agents in production. With AgentControl, teams can change how an agent behaves at runtime without redeploying the underlying application. As AI agents move into production, engineering teams need new ways to manage configuration, quality, and runtime behavior. Unlike traditional code, agent behavior may vary across models, prompts, and production contexts, even when application code remains unchanged. Once deployed, agent configurations … More → The post LaunchDarkly adds real-time controls for AI agents in production appeared first on Help Net Security.
http://news.poseidon-us.com/TScY2X

Canonical ships Ubuntu Core 26 with 15 years of security maintenance

Operators of industrial sensors, edge AI controllers, and connected medical equipment now have a refreshed long-term Linux option for fleets that must stay patched for more than a decade. Canonical released Ubuntu Core 26, the latest long-term supported version of its minimal, immutable operating system, with security maintenance lasting up to 15 years. The release targets devices subject to the European Union’s Cyber Resilience Act and customers running attested edge AI workloads. Every component on … More → The post Canonical ships Ubuntu Core 26 with 15 years of security maintenance appeared first on Help Net Security.
http://news.poseidon-us.com/TScY2V

New macOS infostealer impersonates Apple, Microsoft, and Google in a single attack chain

A SHub macOS infostealer variant called Reaper impersonates Apple, Microsoft, and Google to trick users into executing malicious code, then targets browser data, password managers, and cryptocurrency wallets while establishing persistence for continued access, SentinelOne found. ClickFix gives way to a new delivery method Consistent with earlier SHub versions, Reaper uses a multi-stage execution chain. Researchers said this variant shifts away from standard ClickFix social engineering techniques, where victims are tricked into pasting commands into … More → The post New macOS infostealer impersonates Apple, Microsoft, and Google in a single attack chain appeared first on Help Net Security.
http://news.poseidon-us.com/TScY2R

The end of unencrypted Discord calls is here

Discord has protected voice and video calls in DMs, group DMs, voice channels, and Go Live streams with end-to-end encryption (E2EE) by default. The company began experimenting with E2EE for voice and video in 2023, starting a long-term effort. End-to-end encryption allows only participants in a call to access its content, while Discord does not have access to media encryption keys. Since then, the company introduced DAVE, an open and audited E2EE protocol developed for … More → The post The end of unencrypted Discord calls is here appeared first on Help Net Security.
http://news.poseidon-us.com/TScY1J

The ERP problem that AI was built to solve

For decades, ERP modernization has been too expensive, too slow and too risky to tackle. AI just changed the equation.
http://news.poseidon-us.com/TScXDc

Babel Street targets AI-driven threats with new agentic investigation capabilities

Babel Street has launched Insights Investigator, a new agentic capability that puts tradecraft-trained AI agents at the front edge of investigative work while ensuring analysts remain in control of scope, logic, and outcomes of their missions. As part of the Babel Street Insights platform, Investigator represents a shift from search and AI-assisted queries to analyst-directed, AI-executed investigations. Threat actors are no longer constrained by human bandwidth. Nation-state adversaries, organized criminal networks, and hostile foreign intelligence … More → The post Babel Street targets AI-driven threats with new agentic investigation capabilities appeared first on Help Net Security.
http://news.poseidon-us.com/TScJyk

iProov brings identity verification to video meetings to reduce fraud risks

iProov has launched iProov Verified Meetings, a new solution that enables organizations to verify the identity of video call participants without adding friction to the user experience. Video meetings have become a trusted and scalable communication channel, but attackers are increasingly exploiting them through AI-generated deepfake and injection attacks, creating new fraud risks. Verified Meetings addresses these threats by helping organizations confirm participant identities before meetings begin. The solution is part of the iProov Workforce … More → The post iProov brings identity verification to video meetings to reduce fraud risks appeared first on Help Net Security.
http://news.poseidon-us.com/TScJxq

Egnyte unveils Email Capture and AI features to unify fragmented data

Egnyte has announced a new set of capabilities designed to consolidate fragmented knowledge. Email Capture centralizes critical communications and attachments from siloed inboxes into the Egnyte folder structure, assisting users to make more informed data-driven decisions based on their entire knowledge base. Egnyte is also launching a set of AI-driven integrations and capabilities specifically designed for the architecture, engineering, and construction (AEC) industry. Data fragmentation is a pervasive problem for organizations that can contribute to … More → The post Egnyte unveils Email Capture and AI features to unify fragmented data appeared first on Help Net Security.
http://news.poseidon-us.com/TScFpH

Forget electrons, this breakthrough uses light-matter particles to power AI

Researchers at Penn have created a hybrid light-matter particle that could dramatically speed up AI computing while using far less energy. The breakthrough may help replace some electronic computing processes with ultra-efficient light-based technology.
http://news.poseidon-us.com/TScF8T

Public Instagram posts provide raw material for AI phishing campaigns

A handful of public Instagram posts can give attackers enough material to generate convincing phishing emails with GenAI. Research from the University of Texas at Arlington and Louisiana State University showed how public social media activity can be turned into phishing messages that appear personal and credible to human recipients. Attack pipeline overview (Source: Research paper) The findings highlight a growing problem for security teams and users. Attackers no longer need stolen databases or long … More → The post Public Instagram posts provide raw material for AI phishing campaigns appeared first on Help Net Security.
http://news.poseidon-us.com/TScBgR