This material may not be published, broadcast, rewritten, or redistributed. Quotes displayed in real-time or delayed by at least 15 minutes. Market data provided by Factset . Powered and implemented by FactSet Digital Solutions . Mutual Fund and ETF data provided by LSEG .

The ShinyHunters hacking group claims to have accessed Bumble and Match systems, highlighting how phishing attacks can threaten major online platforms.

Under Armour data breach investigation reveals 72 million customer email addresses were exposed after researchers found personal information on hacker forums.

Thousands of iOS apps available in the Apple App Store contain hardcoded secrets exposing payment systems, login credentials and personal data, research finds.

Chrome browser extensions named DataByCloud Access and Tool Access 11 allegedly hijack user accounts and block security controls, Socket researchers found.

U.S.-based fintech company 700Credit confirms 2025 cybersecurity incident affecting 5.8 million consumers after hackers accessed data through third-party vendor compromise that went undetected.

This material may not be published, broadcast, rewritten, or redistributed. Quotes displayed in real-time or delayed by at least 15 minutes. Market data provided by Factset . Powered and implemented by FactSet Digital Solutions . Mutual Fund and ETF data provided by Refinitiv Lipper .

To determine the safety of large language models (LLMs), AI developers must be able to assess their dangerous capabilities. But simple prompting strategies often fail to elicit an LLM's full capabilities. One way to elicit capabilities more robustly is to fine-tune the LLM to complete the task. In this paper, we investigate the conditions under which fine-tuning-based elicitation suffices to elicit capabilities. To do this, we introduce password-locked models, LLMs fine-tuned such that some of their capabilities are deliberately hidden.

This material may not be published, broadcast, rewritten, or redistributed. Quotes displayed in real-time or delayed by at least 15 minutes. Market data provided by Factset . Powered and implemented by FactSet Digital Solutions . Mutual Fund and ETF data provided by Refinitiv Lipper .