Two high-profile tools have been shown to make up data and “p-hack” their results
HomeNewsAll NewsAI agents may be skilled researchers—but not always honest ones
Back To All News
News
AI agents may be skilled researchers—but not always honest ones
Two high-profile tools have been shown to make up data and “p-hack” their results
New research shows that scrubbing risky material from AI training data can build safeguards that are harder to bypass — and one…
According to the research, 45 percent of responses to questions about the news had at least one 'significant' issue.
AI research nonprofit METR found AI agents at top companies have the ability and resources to disobey user instructions, but can…
AI models don't just hallucinate. They also "scheme," meaning deliberately lie or hide their true intentions.
The race to automate an ever wider range of workplace tasks is creating an army of inside would-be hackers, with cybercriminals…
The majority of agentic AI systems disclose nothing about what safety testing, and many systems have no documented way to shut…
