Generative AI and large language models (LLMs) are revolutionizing natural language processing (NLP), offering enhanced conversational AI experiences for customer service and boosting productivity. To meet enterprise needs, it’s important to ensure the responses that are generated are accurate as well as respect the permissions model associated with the underlying content.

We’re excited to announce that Riscosity has successfully completed its SOC 2 Type II audit. This is a big effort for any organization, and it shows our commitment to protecting our customers’ sensitive data.

Setting and enforcing a policy for strong passwords should be a top priority for organizations in their cyber hygiene practice. Best practices and recommendations keep being updated since this issue is so critical for attack prevention. In fact, compromised passwords are responsible for 81% of successful attacks.

In today's fast-paced business environment, the ever-evolving landscape of technology empowers employees with unprecedented flexibility and agility. While this fosters innovation and productivity, it also presents a lurking challenge—Shadow IT. This term encapsulates the use of unauthorized software, applications, or devices within an organization, posing substantial cybersecurity risks and operational hurdles.

Read also: Sanctions used for the first time for a Russian cybercriminal, Trickbot member gets a 5-year prison sentence, and more.

If you’re using GKE (Google Kubernetes Engine), you should be extremely cautious when adding roles to the system:authenticated group because anyone with a Gmail account can access your cluster.

An authentication bypass vulnerability, tracked as CVE-2024-0204, was discovered in Fortra's GoAnywhere MFT versions prior to 7.4.1 and allows an unauthorized user to create an admin user via the administration portal. This vulnerability has a CVSS score of 9.8 with a high potential for exploitation, which we expect to see in the short term due to a proof of concept (PoC) being available. Fortra informed customers on December 4, 2023, of the flaw via an internal forum post.

In a newly published report, the UK's National Cyber Security Centre (NCSC) has warned that malicious attackers are already taking advantage of artificial intelligence and that the volume and impact of threats - including ransomware - will increase in the next two years. The NCSC, which is part of GCHQ - the UK's intelligence, security and cyber agency, assesses that AI has enabled relatively unskilled hackers to "carry out more effective access and information gathering operations...

A selection of this week’s more interesting vulnerability disclosures and cyber security news. Bad news for some Pixel owners this month – sorry to hear you might have an expensive paper weight now.

Atlassian recently disclosed a new critical vulnerability in its Confluence Server and Data Center product line, the CVE has a CVSS score of 10, and allows an unauthenticated attacker to gain Remote Code Execution (RCE) access on the vulnerable server. There is no workaround, the only solution being to upgrade to the latest patched versions.