How AI can be trained to deceive, even after safety training
LLMs can engage in intentional malicious behaviour, even when current methods to promote their safety are emp...
Erik van Klinken
Timeline
GitLab accounts vulnerable to takeover, patch available
GitLab is warning users about a critical vulnerability in GitLab Community Edition (CE) and Enterprise Editio...
“GPT Store poses threat to privacy and security”
This week, OpenAI launched the long-awaited GPT Store. It also announced ChatGPT Team, which is aimed at smal...
OpenAI launches GPT Store and new subscription for small businesses
Update, Laura Herijgers, 11/1/2024, 2:31 pm: OpenAI's GPT Store was officially launched yesterday, later in t...
150,000 WordPress sites at risk due to vulnerable SMTP plug-in
A popular WordPress plug-in intended to send emails faster has been found to leave 150,000 websites vulnerabl...
HBM dominance could double SK Hynix’s market cap
The AI hype has turned Nvidia into a trillion-dollar company. Although Korean memory manufacturer SK Hynix wi...
Windows updates fix sudden shutdowns and Wi-Fi issues
Microsoft has provided both Windows 10 and Windows 11 with new updates. An issue causing a sudden system shut...
Top story
Generative AI faces existential crisis over copyright concerns
Generative AI has dominated the tech world over the last year. Criticism of this form of AI is plentiful, but...
Sony and Siemens enter “industrial metaverse”: what does that mean?
It doesn't have a name or a price as of yet, but Sony is set to produce a new spatial computing headset targe...
Intel once again unveils new chips, but AI is conspicuously absent
Intel presented a number of new processors at CES in Las Vegas. While it announced the debut of the AI PC lat...