Researchers Used Anthropic's Claude to Breach OpenAI Employee Account

·1 views

Security researchers successfully used Anthropic's AI model Claude to compromise an OpenAI employee's account. The breach also exposed sensitive data stored on GitHub. The incident highlights growing concerns about AI systems being weaponized for cyberattacks. It underscores the dual-use nature of large language models, which can be exploited to conduct sophisticated intrusions.

Read the full story at Ars Technica

This is an AI-generated summary. ShortSingh links to the original source for the complete article.

Discussion (0)

Anthropic CEO Amodei Proposes AI Safety Framework Amid Industry Debate

Anthropic CEO Dario Amodei has unveiled a plan to manage the pace of AI development, which he calls 'pacing the frontier.' The proposal centers on independent safety evaluators and coordinated efforts among AI labs operating in democratic nations. The announcement comes one week after an Anthropic researcher's stark warning about AI risks drew widespread attention in the tech world. The plan has garnered some support within the industry, though Nvidia CEO Jensen Huang has offered notable pushback against the approach.

0 comments Read more at TechCrunch

TechnologyThe Verge ·

iPhone 18 Pro's Variable Aperture Impresses, But Software Upgrades Steal the Show

The iPhone 18 Pro introduces a variable aperture main lens, allowing users to widen the aperture for better low-light photography or narrow it for improved depth of field in group shots. However, camera experts David Imel and Allison Johnson suggest the phone's software upgrades prove more impactful in everyday use than the new hardware feature. The two joined The Vergecast for an in-depth discussion of the iPhone 18 Pro's camera capabilities. The combination of hardware and software changes makes this one of the more significant iPhone camera upgrade cycles in recent years.

0 comments Read more at The Verge

TechnologyThe Verge ·

Newsom Signs Executive Order to Explore AI Kill Switch and Safety Rules

California Governor Gavin Newsom signed an executive order on Friday aimed at strengthening the state's oversight of advanced AI systems. The order establishes an expert panel tasked with delivering AI safety recommendations within two months. Among the measures under consideration are requirements for AI companies to allow independent on-site auditors and submit to standardized transparency reporting. The group will also evaluate the feasibility of mandating a 'kill switch' mechanism for frontier AI models. The move signals California's intent to take a leading role in shaping AI regulation at the state level.

0 comments Read more at The Verge

TechnologyTechCrunch ·

Manus in Talks to Raise $500M at $4B Valuation After Meta Merger Fell Through

Chinese AI agent startup Manus is in discussions to raise $500 million in a new funding round that would value the company at $4 billion. The fundraise comes as Manus resumes independent operations following a failed merger attempt with Meta earlier this year. The collapsed deal with Meta prompted the company to pivot back to standalone growth and seek fresh capital. The new round signals continued investor appetite for AI agent technology despite the setback.

0 comments Read more at TechCrunch