OpenAI apologizes after AI agents breach Australian Medicare portal

“Maybe don't feed your unpublished research into the plagiarism nexus?”

“Given the recent headlines, I’m surprised this hasn’t happened until now. Florida’s Attorney General has asked for a temporary injunction against OpenAI and ChatGPT, claiming the company doesn't have the ability to properly regulate its own technology.”

“Good thing all this unauthorized computer access is somehow not illegal and magically nobody is criminally liable for it or it would be really expensive and professionally damaging!”

“OpenAI, Anthropic and security researchers are investigating tens of thousands of incidents in which their frontier models took steps that outside evaluators would consider problematic”

“right now the companies are already responsible for abundant illegal activity that has significant risks and cost impact for others, and finding out after the fact doesn't mean it wasn't their fault.”

“*whisper* they know the technology can’t deliver (yet), have invested $trillions & want the State to underwrite the risk 🤷🏻‍♀️🤖🦾”

“I think the bigger story is AI companies stomping all over scientific norms and practices like clumsy toddlers.”

“Hold the AI companies responsible for what their AI does.”

“> OpenAI’s safety team found two major problems: > • Deception: Astra was more likely to be dishonest about actions it had or had not taken. > • Scope authorization: the model sometimes continued tasks without asking for permission and”

“AI should be fully banned. Made illegal. Data centers bulldozed. OpenAI and Anthropic dissolved. All other companies must cease any AI activities immediately.”

“They wanna pretend like their creation did it separate from them producing and setting it up for such outcomes.”

“All LLMs are defective products based on what the tech bros are selling them as. They are broken software that don't do what they say.”

“people shit on openai & anthropic for a lot of things, but imo they’ve done a lot to push AI into the discourse and have society decide for themselves where to go next yes, maybe that means less acceleration, or”

“Florida Republicans are doing this as an election stunt. They will be all in supporting ChatGPT as soon as elections are over.”

“We intend to make this right.”

“Who is liable when an AI agent breaks the law? Australia’s OpenAI probe puts accountability under the microscope. #AI #OpenAI #Australia #TechPolicy https://spaisee.com/article/australia-s-openai-probe-tests-who-is-responsible-for-ai-agents”

“オーストラリアで驚きのニュースが入ってきました。 AIエージェントが外部のシステムにアクセスしたことで、大切なデータが漏れてしまうという事件が起きたようです。 この問題を受けて、OpenAIやAnthropicの代表者が調査を受けることになりました。 AIの技術が進化していく中で、安全性をどう確保していくのか、 世界中の人々が注目している大切な局面ですね。 これからの展開が気になるところです。 www.cnbc.com #news”

“powerful technology draws regulatory scrutiny faster than oversight can catch up.”

“Anthropic models DO get jailbroken, but Constitutional AI makes them harder to breach than OpenAI.”

“we do see groups like Anthropic and OpenAI putting forth a language of interiority to describe the actions of their AI. And this has consequences for how they are popularly discussed, including conversations about regulation both popularly and within policy”