‘The machine did it coldly’: Israel used AI to identify 37,000 Hamas targets
Israeli intelligence sources reveal use of ‘Lavender’ system in Gaza war and claim permission given to kill civilians in pursuit of low-ranking militants
AI bots hallucinate software packages and devs download them
Not only that but someone, having spotted this reoccurring hallucination, had turned that made-up dependency into a real one, which was subsequently downloaded and installed thousands of times by developers as a result of the AI's bad advice, we've learned. If the package was laced with actual malware, rather than being a benign test, the results could have been disastrous.
Lass Security's recent research on AI Package Hallucinations extends the attack technique to GPT-3.5-Turbo, GPT-4, Gemini Pro (Bard), and Coral (Cohere).
Microsoft Copilot for Security: General Availability details
Microsoft Copilot for Security will be generally available on April 1st. Read this blog to learn about new productivity research, product capabilities,..
World’s first major act to regulate AI passed by European lawmakers
The European Union Parliament on Wednesday approved the world's first major set of regulatory ground rules to govern the mediatized artificial intelligence at the forefront of tech investment.
Microsoft AI engineer says Copilot Designer creates disturbing images
Shane Jones, who’s worked at Microsoft for six years, has been testing the company’s AI image generator in his free time and told CNBC he is disturbed by his findings. He’s warned Microsoft of the sexual and violent content that the product, Copilot Designer, is creating, but said the company isn’t taking appropriate action. On Wednesday, Jones escalated the matter, sending letters to FTC Chair Lina Khan and to Microsoft’s board, which were viewed by CNBC.
Google CEO Sundar Pichai calls AI tool’s responses ‘completely unacceptable’
In a memo to employees sent Tuesday evening, Sundar Pichai vowed to make structural changes to address the issues found in Gemini’s racially inaccurate images.
New ‘Magic’ Gmail Security Uses AI And Is Here Now, Google Says
Google has confirmed a new security scheme which, it says, will help “secure, empower and advance our collective digital future” using AI. Part of this AI Cyber Defence Initiative includes open-sourcing the new, AI-powered, Magika tool that is already being used to help protect Gmail users from potentially problematic content.
Disrupting malicious uses of AI by state-affiliated threat actors
We terminated accounts associated with state-affiliated threat actors. Our findings show our models offer only limited, incremental capabilities for malicious cybersecurity tasks.
Researchers Say the Deepfake Biden Robocall Was Likely Made With Tools From AI Startup ElevenLabs
Two fake-audio experts say that the deepfake robocall of President Biden received by some voters last week was likely created with technology from Silicon Valley’s favorite voice-cloning startup.
Anthropic researchers find that AI models can be trained to deceive
A study co-authored by researchers at Anthropic finds that AI models can be trained to deceive -- and that this deceptive behavior is difficult to combat.