Here’s an overview of some of last week’s most interesting news, articles, interviews and videos:

Cloudflare moves up its post-quantum deadline as researchers narrow the path to Q-Day
Cloudflare announced it is targeting 2029 to complete post-quantum security across its entire product suite, including post-quantum authentication. The company is following a revised roadmap that Google also adopted after announcing that it had improved the quantum algorithm used to break elliptic curve cryptography. Google stopped short of publishing the algorithm, disclosing only a zero-knowledge proof of its existence.

6G network design puts AI at the center of spectrum, routing, and fault management
Wireless network operators are preparing for a generation of infrastructure where AI is built into the architecture from the start. Sixth-generation networks, expected to reach commercial development over the coming decade, are being designed with AI at the center of how spectrum is allocated, traffic is routed, and failures are detected.

What managing partners should ask AI vendors before signing any contract
In this Help Net Security interview, Kumar Ravi, Chief Security & Resilience Officer at TMF Group, argues that over-privileged access and weak workflow controls pose more danger than ransomware attacks, precisely because they accumulate quietly and go unnoticed.

Chaos malware expands from routers to Linux cloud servers
Chaos, Go-based malware first documented by Lumen’s Black Lotus Labs, has historically targeted routers and edge devices. A new variant observed in March 2026 shows the malware operating against misconfigured Linux cloud servers, a category of infrastructure the botnet had not previously prioritized.

Asqav: Open-source SDK for AI agent governance
AI agents are executing consequential tasks autonomously, often across multiple systems and with little record of what they did or why. Asqav, a Python SDK released under the MIT license, addresses that gap by attaching a cryptographic signature to each agent action and linking entries into a hash chain.

AI agent intent is a starting point, not a security strategy
In this Help Net Security interview, Itamar Apelblat, CEO of Token Security, walks through findings from the company’s research, which shows that 65% of agentic chatbots have never been used yet still hold live access credentials. He explains why organizations treat AI agents more like quick experiments than governed identities, and why that creates risks similar to orphaned service accounts, only harder to see.

Health insurance lead sites sell personal data within seconds of form submission
Lead generation websites that offer health insurance quotes collect sensitive personal data and sell it to multiple buyers within seconds of a user clicking submit. A study by researchers at UC Davis, Stanford University, and Maastricht University mapped this process across 105 health insurance lead generation sites and monitored what happened to the data over 60 days.

What vibe hunting gets right about AI threat hunting, and where it breaks down
In this Help Net Security interview, Aqsa Taylor, Chief Security Evangelist, Exaforce, explains vibe hunting, an AI-driven approach to threat detection that inverts traditional hypothesis-driven methods.Instead of analysts defining attack vectors upfront, the AI scans datasets for anomalous patterns and surfaces potential threats.

Little Snitch for Linux shows what your apps are connecting to
Network monitoring on Linux has long been a gap for users who want per-process visibility into outbound connections. Existing tools either operate at the command line or were designed for server security rather than desktop privacy. Objective Development, the Austrian company behind the macOS firewall utility Little Snitch, released a Linux version of the tool. It is free and, according to the company, will remain so.

To counter cookie theft, Chrome ships device-bound session credentials
Cookie theft follows a well-established pattern. Infostealer malware infiltrates a device, extracts authentication cookies, and exfiltrates them to an attacker-controlled server. Because cookies often have extended lifetimes, attackers can access accounts without passwords, then bundle and sell the stolen credentials. Once malware gains access to a machine, it can read the local files and memory where browsers store authentication cookies.

Social engineering attacks on open source developers are escalating
North Korean hackers spent weeks socially engineering an Axios maintainer through a fake Slack workspace, a cloned company identity, and a fabricated Microsoft Teams call that tricked him into installing a RAT posings as a software update. They used the access they gained to inject malware into npm packages downloaded 100+ million times a week.

BlueHammer: Windows zero-day exploit leaked
A buggy but functional proof-of-concept (PoC) exploit for an unpatched Windows local privilege escalation vulnerability dubbed BlueHammer has been published on GitHub by someone who goes by the handle Chaotic Eclipse and Nightmare Eclipse.

Acrobat Reader zero-day exploited in the wild for many months
Unknown attackers have exploited a zero-day Adobe Acrobat Reader vulnerability since November 2025 and possibly even earlier, security researcher Haifei Li has discovered.

Claude helps researcher dig up decade-old Apache ActiveMQ RCE vulnerability (CVE-2026-34197)
In the latest demonstration of how AI assistants can help with bug hunting, Horizon3.ai researcher Naveen Sunkavally used Claude to unearth CVE-2026-34197, a remote code execution vulnerability in Apache ActiveMQ that’s been introduced in the codebase 13 years ago.

Poisoned “Office 365” search results lead to stolen paychecks
A financially motivated hacking group is targeting Canadian employees with a sophisticated campaign designed to covertly redirect their salary payments into attacker-controlled bank accounts, Microsoft researchers discovered.

ClickFix campaign delivers Mac malware via fake Apple page
Security researchers at Jamf have uncovered a new ClickFix-style attack targeting Mac users via a fake Apple-themed webpage offering instructions on how to “reclaim disk space on your Mac”.

How Mimecast brings enterprise-grade email protection to API deployment
In this Help Net Security video, Andrew Williams, Senior Product Manager at Mimecast, walks through the company’s API-based email security protection for Microsoft 365 and Google Workspace environments.

April 2026 Patch Tuesday forecast: Spring-cleaning of a preview
The first quarter of the year has already passed. Recent discussions on AI and its risks underscore the importance of human oversight. AI can deliver value in many situations, but it can still reach incorrect conclusions even with complete information. For now, the approach remains trust but verify.

IT talent looks the other way as wireless security incidents pile up
Enterprise wireless networks are supporting a growing mix of devices and applications, increasing operational demand and security exposure. The 2026 Cisco State of Wireless report reflects these conditions through rising incident rates, higher costs, and ongoing staffing challenges.

Residential proxies make a mockery of IP-based defenses
Attack traffic moved through ordinary home and mobile connections in ways that limited the usefulness of IP reputation on its own. GreyNoise observed 4 billion malicious sessions during a 90-day period and described activity that appeared indistinguishable from normal user traffic at the network level.

OpenAI opens applications for an external AI safety research fellowship
OpenAI is accepting applications for a paid fellowship program that will fund external researchers to work on safety and alignment questions related to advanced AI systems. The program, called the OpenAI Safety Fellowship, runs from September 14, 2026 through February 5, 2027. Applications close May 3, with successful applicants notified by July 25.

Russian hackers hijack internet traffic using vulnerable routers
The Russian state cyber group APT28 has been compromising routers to hijack web traffic and spy on victims, the UK’s The National Cyber Security Centre (NCSC) has warned. Attackers are exploiting vulnerable routers to alter DHCP and DNS settings, redirecting traffic through servers they control.

Cybercrime losses break the $20 billion mark
Online crime continues to generate rising financial losses, with totals reaching $20.877 billion in 2025. The FBI’s Internet Crime Complaint Center (IC3) report shows a 26% increase in total reported losses from the previous year. More than one million complaints were submitted during the year, with fraud accounting for the majority of losses. Cyber-enabled fraud totaled $17.7 billion, representing 85% of all reported financial damage.

Cybercriminals move deeper into networks, hiding in edge infrastructure
Attack activity is moving toward infrastructure outside endpoint visibility. Proxy networks support a wide range of operations, edge devices serve as initial access points, and GenAI speeds up how attackers assemble and rebuild their tooling. Lumen’s 2026 Threatscape Report describes this pattern in criminal and nation-state activity.

OpenSSL 3.6.2 lands with eight CVE fixes
OpenSSL 3.6.2 patches eight CVEs across a range of components. The project rates the most severe issue in the release as Moderate. The release fixes incorrect failure handling in RSA KEM RSASVE encapsulation (CVE-2026-31790) and a loss of key agreement group tuple structure when the DEFAULT keyword is used in server-side configuration of the key-agreement group list (CVE-2026-2673).

Prompt injection tags along as GenAI enters daily government use
Routine use of GenAI has moved into daily operations in state and territorial government environments, placing new security risks within common workflows. A Center for Internet Security (CIS) report, Prompt Injections: The Inherent Threat to Generative AI, identifies prompt injection as a persistent concern tied to that adoption.

Phishers sneak through using GitHub and Jira’s own mail delivery infrastructure
Attackers are abusing the notification systems of SaaS platforms like GitHub and Jira to send phishing and spam emails, Cisco Talos researchers are warning. On GitHub, the attackers are abusing the service’s notification system to deliver malicious payloads.

WhatsApp brings long-awaited privacy feature to filter who can reach you
After years of waiting, WhatsApp is set to roll out a username feature that will allow people to connect and communicate without sharing their phone numbers. This means more privacy and better control over phone number visibility by choosing a unique username.

113,000 explicit prompts from AI girlfriend platform exposed, many linked to user IDs
MyLovely.AI, an AI girlfriend platform, suffered a data breach that exposed over 100,000 users. MyLovely.AI allows people to create personalized not safe for work (NSFW) content and engage in real-time conversations with AI-generated companions, often involving highly personal prompts and interactions.

CISOs grapple with AI demands within flat budgets
Security spending continues to edge upward across large organizations, though the changes remain gradual and tightly managed. The 2026 RH-ISAC CISO Benchmark reflects a steady environment where budgets expand in small steps, even as AI becomes a routine part of security operations.

Product showcase: Proton Authenticator is an end-to-end encrypted, open source 2FA app
Proton Authenticator is a free and open-source two-factor authentication (2FA) app that generates time-based one-time passwords (TOTP) to help secure online accounts. It is available on Windows, macOS, Linux, iOS, and Android, allowing users to access their verification codes across devices.

Google study finds LLMs are embedded at every stage of abuse detection
Online platforms are running large language models at every stage of LLM content moderation, from generating training data to auditing their own systems for bias. Researchers at Google mapped how this is happening across what the authors call the Abuse Detection Lifecycle, a four-stage framework covering labeling, detection, review and appeals, and auditing.

Comp AI: The open-source way to get compliant with SOC 2, ISO 27001, HIPAA and GDPR
Getting a startup through a SOC 2 audit has long meant months of manual evidence collection, policy writing, and repeated back-and-forth with auditors. A growing number of compliance platforms have moved to automate parts of that process, and Comp AI is now doing it with an open-source codebase that organizations can inspect, modify, and self-host.

GitHub Copilot CLI gets a second-opinion feature built on cross-model review
Coding agents make decisions in sequence: a plan is drafted, implemented, then tested. Any error introduced early compounds as subsequent steps build on the same flawed assumption. Self-reflection is a recognized mitigation technique, and one GitHub Copilot already supports, but a model reviewing its own output is still constrained by the same training data and blind spots that produced it.

AI-enabled device code phishing campaign exploits OAuth flow for account takeover
A phishing campaign that bypasses the standard 15-minute expiration window through automation and dynamic code generation, leveraging the OAuth Device Code Authentication flow to compromise organizational accounts at scale, has been observed by the Microsoft Defender Security Research team. The campaign uses AI-assisted infrastructure and end-to-end automation.

Anthropic’s new AI model finds and exploits zero-days across every major OS and browser
Automated vulnerability discovery tools have existed for decades, and the gap between finding a bug and building a working exploit has always slowed attackers. That gap is now substantially narrower. Anthropic’s Claude Mythos Preview, a new general-purpose language model being made available only to a limited group of critical industry partners and open source developers, can autonomously identify zero-day vulnerabilities and then construct working exploits across every major operating system and major web browser.

Flatpak 1.16.4 fixes sandbox escape and three other security flaws
Flatpak, a Linux application sandboxing and distribution framework, released version 1.16.4, patching four security vulnerabilities. The most severe fix addresses a complete sandbox escape that leads to host file access and code execution in the host context, tracked as CVE-2026-34078.

Iranian cyber activity hits US energy, water, and government networks
U.S. government agencies on Tuesday warned American organizations about ongoing cyber activity targeting OT and PLC devices, including those manufactured by Rockwell Automation and Allen-Bradley, across multiple critical infrastructure sectors. The activity has been attributed to Iranian-affiliated APT actors seeking to disrupt operations in the United States.

Meta’s Muse Spark takes AI a step closer to personal superintelligence
Meta Superintelligence Labs has introduced Muse Spark, a natively multimodal reasoning model with support for tool use, visual chain of thought, and multi-agent orchestration. The release includes a Contemplating mode, which is rolling out gradually and orchestrates multiple agents that reason in parallel.

Claude Managed Agents bring execution and control to AI agent workflows
Anthropic’s Claude Managed Agents are a suite of composable APIs for building and deploying cloud-hosted agents at scale, handling sandboxed code execution, checkpointing, credential management, scoped permissions, and end-to-end tracing for you.

Product showcase: Session, a messenger without phone numbers or metadata
Instant messaging has been around for decades, but it became widely adopted with the emergence of smartphones. Earlier, communication was limited to basic text messages. Messaging expanded to include photos, videos, and video calls without relying on telecom networks, as long as there is a reliable data connection.

Gmail’s end-to-end encryption comes to mobile, no extra apps required
Google has expanded Gmail client-side encryption to Android and iOS devices, allowing users to engage with their organization’s most sensitive data on mobile devices while ensuring data remains compliant with sovereignty and compliance requirements. This feature is available for Enterprise Plus users with the Assured Controls or Assured Controls Plus add-on.

Cybersecurity jobs available right now: April 8, 2026
We’ve scoured the market to bring you a selection of roles that span various skill levels within the cybersecurity field. Check out this weekly selection of cybersecurity jobs available right now.

New infosec products of the week: April 10, 2026
Here’s a look at the most interesting products from the past week, featuring releases from Advenica, Intruder, Mallory, and Secureframe.