The latest Threat Detection Report update from Red Canary shows a rise of almost 500 percent in detections associated with cloud accounts during the first half of 2025.

This significant rise stems primarily from Red Canary’s expanded identity detection coverage and the implementation of AI agents designed to identify unusual login patterns and suspicious user behaviors. This includes identifying logins from unusual devices, IP addresses, and virtual private networks (VPNs), which significantly increases the detection of risky behaviors.

A disconnect between cybersecurity confidence and data reality is leaving organizations exposed, according to a new report released today by Axonius.

The study, based on a survey of 500 US director-level and above cybersecurity and IT leaders, reveals that while 90 percent of cybersecurity leaders say their organization is prepared to take immediate action on a vulnerability, only 25 percent trust all the data in their own security tools.

New analysis from Forescout of more than 23,000 vulnerabilities and 885 threat actors across 159 countries worldwide during the first half of 2025 finds 47 percent of newly exploited vulnerabilities were originally published before 2025, and zero-day exploitation has increased 46 percent.

The report also shows ransomware attacks are averaging 20 incidents per day, zero-day exploits increased 46 percent, and attackers are increasingly targeting non-traditional equipment, such as edge devices, IP cameras and BSD servers. These footholds are often used for lateral movement across IT, OT, and IoT environments, allowing threat actors to get deeper into networks and compromise critical systems.

Adversaries are weaponizing GenAI to scale operations and accelerate cyberattacks -- as well as increasingly targeting the autonomous AI agents reshaping enterprise operations. This is among the findings of CrowdStrike’s 2025 Threat Hunting Report.

The report reveals how threat actors are targeting tools used to build AI agents -- gaining access, stealing credentials, and deploying malware -- a clear sign that autonomous systems and machine identities have become a key part of the enterprise attack surface.

Modern IT environments are massively distributed, cloud-native, and constantly shifting. But traditional monitoring and AIOps tools rely heavily on fixed rules or siloed models -- they can flag anomalies or correlate alerts, but they don’t understand why something is happening or what to do next.

We spoke to Casey Kindiger, founder and CEO of Grokstream, to discuss new solutions that blend predictive, causal, and generative AI to offer innovative self-healing capabilities to enterprises.

As we reported earlier this week, the UK’s new Online Safety Act has seen a surge in interest in the use of VPNs and an online petition for its repeal has been signed by over 400,000 people.

An article published yesterday by The Critic argues that the legislation is badly drafted. Industry figures too are raising doubts about the effectiveness of the act, its likely wider impact on cybersecurity and its potential for overreach.

Historically test and measurement has been simply about collecting data and exporting it for later analysis. Now though neural networks make it possible to carry out the analysis in real time.

We spoke to Daniel Shaddock, CEO of Liquid Instruments, to find out more about what this means for businesses.

According to new research from Radware 83 percent of credential stuffing campaigns include explicit API-targeting techniques.

The report shows a shift in credential stuffing attacks, underscoring a fundamental transformation from volume-based attacks leveraging a series of repeated password attempts to more sophisticated, multi-stage infiltration techniques.

As organizations embrace digital transformation and AI, security teams face mounting pressure to defend an ever-expanding attack surface according to a new report.

The research from Cobalt suggests traditional reactive security measures cannot keep pace with modern threats, particularly when adversaries leverage automation and AI to scale their attacks. 60 percent of respondents believe attackers are evolving too quickly for them to maintain a truly resilient security posture.

A new report from Semperis, based on a study of almost 1,500 organizations globally, shows that hackers are stepping up threat levels and ransomware is still a global epidemic.

In 40 percent of attacks threat actors threatened to physically harm executives at organizations that declined to pay a ransom demand. US-based companies experienced physical threats 46 percent of the time, while 44 percent of German firms experienced similar forms of intimidation.

A new report looks at the state of post-quantum cryptography (PQC) from the perspective of
cybersecurity professionals, finding that 48 percent of organizations aren’t prepared to confront the urgent challenges posed by quantum computing.

The report from Keyfactor, based on a survey of 450 cybersecurity leaders across North America and Europe carried out by Wakefield Research, finds mid-sized organizations are particularly vulnerable, with 56 percent saying they are not ready.

Security operations teams often struggle with complex tools, legacy pattern-matching DLP, manual policy tuning, and alert fatigue. This can slow investigations, increase overhead, and reduce security effectiveness.

While traditional DLP solutions aim to tackle these challenges, they require constant human intervention, generate high false positive rates, and often miss sophisticated threats that bypass simple pattern recognition. That’s why Nightfall is launching an autonomous Data Loss Prevention platform.

The UK’s new Online Safety Act came into force this week, aimed at protecting youngsters with age verification to access adult and harmful content. However, it’s seen other material being blocked and sparked concern among free speech campaigners about government censorship.

It’s not too surprising then that there’s been a lot of interest in VPNs since the act came into force. VPNMentor has seen a 6,430 percent peak surge in VPN demand since the act’s introduction.

Cloud threats are evolving faster than most security teams can respond, and traditional security tools are struggling to keep pace. According to IBM’s 2024 Cost of a Data Breach Report it now takes an average of 258 days to detect and contain a breach -- giving attackers more than enough time to access sensitive data and move laterally through cloud infrastructure undetected.

We spoke to CEO of Upwind, Amiram Schacha, to learn why organizations need real-time visibility and protection at the runtime layer -- where threats actually occur -- in order to close this growing security gap.

Electronic data interchange (EDI) is the lifeblood of modern business, but even a small error -- be it a connection failure, data quality issue, transformation failure, or data transmission issue for example -- can rapidly cascade, generating hundreds or even thousands of issues.

This can become a domino effect tipping over into longer root cause identification, inefficiency in managing a raft of open tickets, and a prolonged time to resolution. These factors can increase operational risk, leading to downstream supply chain issues that can jeopardize valuable business relationships.