Author: Ian Vielma

Last Christmas, it was reported that people in the UK were scammed out of more than £11.5 million. Richard Horne, a cybersecurity chief with the UK’s NCSC (National Cyber Security Centre), stated that scammers “targeted bargain hunters with increasingly sophisticated scams, sometimes crafted using AI, making them harder to detect.” Between November 2023 and January 2024, there were more than 16,000 reports of online shopping fraud, with victims losing £695 on average. With Black Friday coming up, the NCSC hopes to spread public awareness educating consumers on ways to prevent getting scammed Full Story

Microsoft has reported that there are over 600 million cyberattacks occurring daily, with cybercriminals using tactics such as password spraying, breach replays, and brute force to commit these acts. These tactics exploit users who use weak passwords or repeated passwords across platforms. In addition, ransomware has remained one of the most prominent cybersecurity concerns over the years; it has continuously evolved over time and isn’t going anywhere. It is recommended that companies use AI-powered tools to detect threats and anticipate potential attacks. Full Story

OpenAI’s Whisper transcription tool has been criticized for producing “hallucinations,” or fabricated text, which can include harmful content and inaccuracies, especially concerning medical transcriptions and captions for the Deaf and hard of hearing. Despite warnings from OpenAI against using Whisper in high-risk areas like healthcare, many medical centers have adopted it to transcribe patient consultations, raising concerns about patient safety and accuracy. Experts are calling for regulatory measures to address these issues, emphasizing the need for improvements in Whisper’s accuracy and the potential dangers of reliance on AI in critical decision-making contexts. Full Story

Researchers tested OpenAI’s GPT-4 at the University of Colorado Health from September 2023 to March 2024. They concluded that the AI-generated responses in patient portal messages improved workflow. Most notably for nurses, allowing them to reply faster, express empathy better, and reduce the need to forward messages. Medical assistants’ and clinicians’ workflow, however, were less impacted. Nurses were more likely to recommend the chatbot, with 92% reporting faster replies, compared to lower percentages among medical assistants and clinicians. The varying types of messages received by nurses, medical assistants, and clinicians may explain the differing views on AI, with the chatbot…

In a survey of more than 3000 public servants, 75% of Europeans support AI technologies, such as facial recognition and biometric data, for police and military surveillance despite Europe’s strict privacy laws. While AI has high approval for public services like traffic optimization, 64% of Europeans oppose its use in decisions like parole. And, of the surveyed, 67% fear AI manipulation in elections, with deepfakes and misinformation posing significant threats to the process. Full Story

A survey by ISACA reveals that 45% of organizations exclude cybersecurity teams from the development, onboarding, and implementation of AI solutions despite the growing use of AI in security operations. Only 35% of cybersecurity professionals participate in creating policies governing AI usage, highlighting a gap in oversight and governance as AI becomes more integral to security strategies. Experts emphasize that cybersecurity teams must be involved in AI solution development to effectively address evolving threats, especially as reliance on AI and automation increases amid staffing shortages in the industry. Full Story

OpenAI recently reported the activities of three hacker groups, CyberAv3ngers and Storm-0817 (both based in Iran), and SweetSpectre (based in China). They are said to have exploited ChatGPT for cyberattacks, reconnaissance, and malware development. The CyberAv3ngers targeted industrial control systems in the U.S. and Ireland, explicitly focusing on poor security practices such as exposed systems and default credentials. The group Storm-0817 used ChatGPT for malware development aimed at Android devices and for scraping social media. At the same time, SweetSpectre attempted to use it for social engineering and delivering malware to OpenAI employees, although those attempts were blocked. Full Story 

The Comcast Business 2024 Cybersecurity Threat Report highlighted a significant rise in cyber threats, driven by cybercriminals leveraging AI for tactics like phishing and exploiting vulnerable systems, with over 2.6 billion phishing interactions detected. Despite these challenges, Comcast believes companies can improve resilience by adopting a combination of advanced protection, detection methods, managed services, and proactive security practices to combat evolving cyber threats. Full Story

Hybrid AI combines various AI techniques and models to leverage their strengths and mitigate weaknesses, enabling more accurate and reliable outcomes across industries, such as healthcare and finance. This approach improves the transparency of AI decision-making, allowing for better understanding and trust, particularly in regulated sectors. While Hybrid AI offers significant promise for developing more powerful and ethical AI systems, it also faces challenges like integration complexities and resource demands. Full Story

Microsoft has been introducing new features to their AI, Microsoft Copilot.  They have pushed out a slightly modified interface and a new Copilot Voice feature for conversational interactions. Copilot Labs will introduce new capabilities like Copilot Vision for interpreting web content and Think Deeper for complex queries, only available for Copilot Pro Subscribers. Features like Copilot Daily for personalized updates will be gradually released, with privacy measures ensuring that data from Copilot Vision sessions are not stored. Full Story