Evaluating LLM safety, bias and accuracy [Q&A]
Large language models (LLMs) are making their way into more and more areas of our lives. But although they're improving all the time they're still far from perfect and can produce some unpredictable results.
We spoke to CEO of Patronus AI Anand Kannappan to discuss how businesses can adopt LLMs safely and avoid the pitfalls.
Public sector and infrastructure come under attack as malicious web requests rise
The number of malicious web requests rose by 53.2 percent in the first half of 2024, compared to the same period last year according to a new study.
The report from German cybersecurity company Myra finds that for the first quarter of 2024, the number of malicious requests on websites, online portals and web APIs increased by 29.8 percent compared to 2023. In the second quarter, the growth was even more pronounced at 80 percent.
Uncovering GenAI's unsung heroes [Q&A]
There's no doubt that AI is seen as flavor of the month across many sectors at the moment. But how much of this is hype and how much genuine value?
We spoke to Martin Hawksey, collaboration engineer at Qodea, to discuss GenAI and the areas where GenAI is making a real difference, some of which you may not be aware of.
Severity of ransomware attacks increases by 68 percent
A new report from insurance provider Coalition finds that that ransomware claims severity spiked by 68 percent in the first half of 2024 to an average loss of $353,000.
While high ransomware demands have come back into vogue, funds transfer fraud (FTF) has also seen a notable decrease in both frequency (two percent) and severity (15 percent).
Qualys launches cloud-based risk operations center
Organizations are facing an ever-growing volume of risk alerts spread across multiple, disconnected top-10 dashboards. This fragmented view can result in conflicting analyses, duplicate work, missed threats, and strategies that fail to fully protect the organization.
Cloud-based security firm Qualys is addressing this with the launch of a new Risk Operations Center (ROC) with Enterprise TruRisk Management (ETM) that enables CISOs and business leaders to manage cybersecurity risks in real time, transforming fragmented, siloed data into actionable insights that align cyber risk operations with business priorities.
Is the IT enterprise helpdesk an endangered species?
The IT helpdesk could be extinct by 2027 according to a new survey of 1,000 global IT workers which finds that 79 percent believe the current service desk model will be unrecognizable within three years, with 77 percent saying new technologies will render it 'redundant' by 2027.
The study by digital employee experience company Nexthink shows 87 percent of IT workers also report that, with digital transformation having greatly expanded the size and complexity of enterprise IT environments, incident response is 'economically unsustainable' unless helpdesks have significant proactive capabilities.
Google launches new initiative for sharing scam details
Google has announced a new partnership with the Global Anti-Scam Alliance (GASA) and the DNS Research Federation (DNS RF) to launch Global Signal Exchange (GSE), a new project with the ambition to be a global clearinghouse for online scams and fraud bad actor signals.
This collaboration combines the strengths of each partner: GASA's extensive network of stakeholders, the DNS Research Federation's robust data platform with already over 40 million signals, and Google's experience in combating scams and fraud.
Attacks on GenAI see sensitive data leaked
New research reveals that 90 percent of successful attacks against GenAI have resulted in the leakage of sensitive data.
The report from Pillar Security, based on real-world analysis of more than 2,000 AI applications, shows 20 percent of jailbreak attack attempts successfully bypassed GenAI application guardrails and adversaries needed an average of just 42 seconds to execute an attack.
AI boosts rise in phishing and spoofing attacks on banks
The US banking industry has seen a significant uptick in cyberattacks, particularly in phishing and spoofing, and tactics are becoming increasingly advanced due to AI.
New research from BforeAI analyzed 62,074 domains registered between January and June 2024 with finance-related keywords. Of those registered domains, 62 percent were found to be involved in phishing attacks targeting legitimate entities via spoofing websites.
Security awareness training is not enough to stop breaches
New research shows companies are pouring an increasing number of resources into their security awareness and training programs, with 96 percent of respondents allocating it between five percent to 20 percent of their security budgets.
But the research from CultureAI, based on a survey by Opinion Matters of 200 UK-based cyber security teams at organizations with over 1000 employees, finds that while 78 percent train employees at least monthly human-related breaches are still happening at an alarming rate.
Why CISOs need to keep on top of their cybersecurity investments [Q&A]
With the current climate of new and more sophisticated cybersecurity products, greater risk and pressure on budgets, it's more important than ever to ensure that they get the best return on their investments.
We spoke to Karthik Swarnam, chief security and trust officer at ArmorCode, to discuss the best practices to ensure ROI from security investments.
'Toxic cloud triad' puts four in 10 organizations at risk
It might sound like a Chinese secret society, but a new report warns that a 'toxic cloud triad' of publicly exposed, critically vulnerable and highly privileged cloud workloads is putting almost four in 10 organizations at risk.
Security gaps caused by misconfigurations, risky entitlements and vulnerabilities combine to dramatically increase cloud risk according to the Tenable Cloud Risk Report.
Enterprise CISOs worry about losing their job after a breach
A new survey, which polled 200 CISOs from companies with annual revenues exceeding $500 million, highlights growing concerns across a number of areas.
The study from Portnox finds worries around the effectiveness of zero trust, the limitations of multi-factor authentication (MFA), and a looming threat to job security amidst an increasingly complex cybersecurity landscape.
New tool helps enterprises identify credential risk
Though organizations and their users have greatly benefited from hybrid environments and SaaS applications, dispersed workforces have also created unforeseen risks that are increasingly difficult to secure.
The average enterprise user has more than 50 passwords not protected by single sign-on (SSO), each representing a point of risk if not properly monitored and secured.
Email attacks target the healthcare sector
A new report from Abnormal Security reveals a rise in targeted email compromise attacks on the healthcare sector.
Vendor email compromise (VEC) attacks on the sector have consistently trended upward, recording a 60 percent increase between August 2023 and August 2024. The sector's reliance on long-term vendor relationships is being exploited through VEC, where cybercriminals impersonate trusted vendors to bypass traditional email security and trick employees.
Ian's Bio
Ian spent almost 20 years working with computers before he discovered that writing about them was easier than fixing them. Since then he's written for a number of computer magazines and is a former editor of PC Utilities. Follow him on Mastodon
© 1998-2025 BetaNews, Inc. All Rights Reserved. Privacy Policy - Cookie Policy.