00:00Before watching the video, remember to check out trending websites for more content!
00:04People assume that the content of your private chats with AI is absolutely...
00:09hundred percent
00:09Is it 100% confidential?
00:11Not a single word will be leaked?
00:13The answer is not necessarily.
00:14Of course, we're not saying that AI-powered chat platforms are untrustworthy.
00:16It's just that they have their own policies.
00:18There was a woman in Florida, USA...
00:21Recently, AI assistants
00:23It's like Claude treated it as her private diary.
00:26It's about venting her various emotions.
00:28And then in the dialogue...
00:30There was mention of plans to attack the police station.
00:35And what about the AI's parent company?
00:36This was determined to be a potential threat.
00:38I called the police immediately.
00:40What about this woman?
00:41He was subsequently arrested by law enforcement officers at his home.
00:43He is currently facing extremely serious felony charges.
00:47He could face up to 15 years in prison.
00:50According to the police...
00:52This 40-year-old woman, Carrie
00:54I usually use Entropic's AI assistant, Claude.
00:57Used as her diary
00:59Record her daily life
01:00There are also some negative emotions.
01:02However, late at night on September 26th
01:04She lost control of her emotions
01:05During the dialogue, it was claimed that
01:07Go to the local sheriff's office
01:09Go and carry out strafing
01:10She then added...
01:12He got a new gun
01:14They even said this was the last chance.
01:16I'm done
01:17So these radical remarks
01:19Claude's security detection mechanism was triggered instantly.
01:22The conversation was also reported to law enforcement agencies.
01:26After manual review?
01:28Entropic's official stance is...
01:30In that case...
01:31It already constitutes a credible threat of violence.
01:34so what
01:34Just record the relevant conversations.
01:36Notify local law enforcement agencies
01:38After the police confirmed Carrie's identity and address, what happened next?
01:42A few days later, they came to arrest him.
01:44During the investigation, Carrie
01:47She then defended herself.
01:48She said she was just venting her emotions on the AI.
01:51She has no intention of taking any concrete action.
01:54However, the police still acted in accordance with Florida law.
01:57Using threats to commit terrorist acts
01:59and second-degree felony of mass shooting
02:02To sue this woman
02:04According to Entropic's official policy
02:06If the company detects a potential emergency
02:09It is possible to bypass the usual legal procedures.
02:11Disclosing user information to the government
02:14Avoiding risks
02:15Even if the user turns off model training
02:18There is still a possibility that conversations involving security issues may be subject to censorship.
02:22This rare case
02:24This also serves as a safety warning for the vast number of AI users.
02:26AI is not an unregulated private diary.
02:31We still need to be vigilant about every word and action in the online world.
02:34To bear legal responsibility
02:35The woman is currently only charged.
02:37He has not yet been convicted.
02:38As for the dialogue in the case
02:40Does the legal provision of the charge apply?
02:42Can the prosecution prove that it truly constitutes a crime, etc.?
02:46Further instructions are needed.
02:48The judicial process will determine the outcome.
02:50Please feel free to like, subscribe, share, and tip to support the Mingjing & DianDian column.
02:53Please feel free to like, subscribe, share, and tip to support the Mingjing & DianDian column.
