DataBreachToday.com

Most Compromises Trace to Financial Services, Healthcare, Professional Services
Data breaches rage on. In the first half of this year, the Identity Theft Resource Center counted 1,732 total data breaches affecting 166 million people, marking a rise in data breaches but a decline in victims, likely due to a drop in mega-breaches.

How Focused Skill Building Solves Real Problems in Cyber Roles
The pressure to grow doesn't come from curiosity alone. It comes from real friction in the systems you work with. That's why the smartest way to continue learning is not to try to master everything. Instead, focus on the next thing that will actually help you move forward in your role.

Experts Aim to Probe How AI Models Reason, and Why It Matters
AI researchers from OpenAI, Google DeepMind and Anthropic and others have urged deeper study into chain-of-thought monitoring, a technique to track how reasoning models arrive at answers. Their joint paper warns that transparency may erode if not prioritized.

5G OT Security Summit Speakers on Delicate Balance Between Innovation, Cyber Risk
Digital transformation - which now includes a convergence of cloud-based applications, AI and OT systems - introduces new threat vectors particularly as legacy systems struggle to adapt. Speakers at the 5G OT Security Summit discussed cyber defenses and policies and for securing OT systems.

67 Malicious Packages, XORIndex Loader Target JavaScript Code-Sharing Platform
North Korean threat actors escalated their software supply chain attacks by uploading 67 new malicious packages to the npm Registry as part of the ongoing Contagious Interview campaign. The malware targets open-source JavaScript developers with malware loaders.

Agency to Collaborate with External Experts on Vulnerability Research
The U.K. NCSC will collaborate with industry experts for vulnerability detection and mitigation as part of its latest Vulnerability Research Initiative. The announcement comes on the heels of funding concerns for the U.S. government-based Common Vulnerabilities and Exposures program.

Security Leaders Need Deep Observability to Balance Innovation and Risk
Organizations face mounting pressure to accelerate AI adoption while maintaining robust security controls across hybrid cloud environments where traditional tools fall short. This World AI Appreciation Day, it's time to challenge the assumption that rapid innovation comes at the cost of security.

Startup Raises $30M, Uses Risk Intelligence to Preempt Reconnaissance Attacks
Former FireEye and Mandiant leader John Watters unveils iCounter, a new cyber risk intelligence startup focused on targeted attacks and AI-enabled adversaries. Backed by Syn Ventures, the firm aims to transform threat detection with deeper visibility into attacker reconnaissance.

Oracle Cloud Infrastructure Flaw Enabled Malicious File Uploads, Researchers Found
Exploring Oracle Cloud Infrastructure, researchers at Tenable found that Oracle's console-based Code Editor tool failed to block arbitrary file uploads, and could be silently exploited via drive-by attacks to install malware. They said Oracle has now fixed the vulnerability.

Hacking Group UNC6148 Steals Credentials With New OVERSTEP Rootkit, Google Says
A cybercrime group used a backdoor in a fully patched SonicWall appliance to steal credentials and may have sold the stolen data to ransomware groups as part of an ongoing campaign, Google Threat Intelligence Group found. The firm attributed the campaign to a cybercrime group it tracks as UNC6148.

How OpenAI Lost, Google Hedged and Cognition AI Bought AI Coding Startup Windsurf
As with any story about AI start-ups, the acquisition of Windsurf AI happened at lightning speed. In just a couple of weeks, major Silicon Valley players took a close look at the four-year-old AI-assisted coding vendor, and ultimately Google and Cognition AI ended up with parts of the company.

Analysts Warn White House Chip Reversal Threatens US AI Dominance
President Donald Trump has reportedly reversed a U.S. policy restricting Nvidia chips critical to artificial intelligence technology development from being exported to China in a move that experts say could undercut the U.S. lead in the burgeoning emerging technology market.

Escalating Geopolitical Tensions Could Increase Hacks by 2030, Government Warns
France has identified Russia as the primary threat to national security in the coming years. The French government recommends adopting measures to strengthen its cybersecurity defenses in anticipation of increased hacks from Moscow-aligned attackers driven by geopolitical tensions.

The flood of new artificial intelligence tools, including those to help cybersecurity teams, can overwhelm healthcare CISOs and their security staff, fueling "AI fatigue" that in itself can create additional cyber risk, said Drew Henderson and Jon Hilton, practice leaders at consulting firm LBMC.

5G OT Security Summit Speakers on Secure Frameworks for Regional Infrastructure
At a time when ASEAN nations are accelerating 5G deployments, cybersecurity leaders at the 5G and OT Security Summit in Malaysia issued a sobering warning: Fragmented regulations and uneven OT readiness threaten to undermine the region’s digital ambitions.

AI Coding Assistants May Hinder, Not Help, Experienced Developers
If AI coding tools were meant to be the espresso shot of developer productivity, a recent study suggests they might actually be more like decaf - offering the aroma of efficiency without the buzz, based on findings from a nonprofit AI research group.

Hacker Claims to Have Exploited Flaw in Oracle WebLogic Server, Sold Stolen Data
Seychelles Commercial Bank is warning customers that a hacker stole their personal information - but no money - from their accounts after breaching its systems. The hacker involved claims to have stolen and sold two gigabytes of customer data from the bank, which paid no ransom.

Attackers Can Upload Malware in System Management Mode and Disable Secure Boot
Multiple high-severity vulnerabilities in Gigabyte's UEFI firmware could enable attackers to execute arbitrary code within System Management Mode, granting persistent and nearly undetectable control over affected systems, cybersecurity researchers reported.

Comprehensive Machine Identity Security Needed for Non-Human Identities
A study from CyberArk shows that machine identity-related security incidents are increasing as the volume and complexity of machine identities surge. Security leaders must build an end-to-end strategy to secure non-human identities and prevent attacks and outages.