Exchange Ranking
Top Exchanges
24h Volume Ranking
Popularity Ranking
Exchange BTC Balance
Proof of Reserves
Decentralized Exchanges
Funding Rate
Funding Heatmap
Liquidation Data
Max Pain
Long/Short Ratio
Whale L/S Ratio
Binance/Okex/Huobi L/S
Bitfinex Margin L/S
ETF Tracker
Solana ETF
XRP ETF
Hong Kong ETF
Bitcoin Treasuries
Crypto Reversal
Ethereum Reserves
HyperLiquid Wallet Analysis
Hyperliquid Whale Watch
Large Transactions
On-chain Movement
Bitcoin ROI
Stablecoin Market Cap
Options Analysis
News
Articles
Economic Calendar
Features
Wallet
Contract Calculator
Security
Collections
Watchlist
FollowingFlorida woman uses AI as a diary but faces felony charges for "threatening remarks"
A Florida woman said she uses artificial intelligence like she writes a diary. Now, however, she is facing a felony charge and the "reader" sent her to prison was the same person she never intended to show content to.
According to local WINK News, Bonita Springs resident Carli Michelle Heller was arrested. Investigators found that she wrote in a Sept. 26 record that she would "shoot" the Lee County Sheriff's Office. Investigation revealed that her Claude account showed that she mentioned in a conversation the next day that she had obtained a new gun.
The arrest report states that Anthropic's platform monitors key phrases and threatening content. Given the seriousness of the statement, the system forwarded it to the manual review team, which then reported it to law enforcement. The police confirmed Heller's identity, went to his residence and detained him without conflict. She is now charged with "making written threats of violence."
Anthropic's relevant documents did not conceal this. Its consumer clause reserves the right to report user input, output or behavior to law enforcement at the company's "absolute discretion." Its privacy policy, effective September 10, 2026, also stipulates that data can be shared with police when companies believe in good faith that disclosure is reasonably necessary to prevent serious harm.
实际上，即使选择退出将其数据用于模型训练，Anthropic仍有权在三个月内（而非五年）使用您的数据，这意味着您的聊天内容并非完全私密。
该政策还指出，被标记为安全审查的对话仍会被用于提升Anthropic检测有害内容的能力。因此，一篇被标记的日记条目最终可能成为检测器的训练材料。
类似事件并非孤例
海勒并非首例。《旧金山标准报》（The San Francisco Standard）今年九月报导称，Anthropic曾在一位用户涉嫌撰写其购买了AR-15步枪并将CEO达里奥·阿莫迪（Dario Amodei）置于瞄准镜中后，将该用户移交警方。一名员工告诉警官，他当时并未立即感到安全受到威胁，并援引公司政策拒绝展示相关消息。
该用户告诉《标准报》，他只是在开玩笑，且在故事发布时并未被捕或受到指控。Anthropic的一位发言人称，这一事件表明“我们的安全机制正在按预期运作”。该公司拒绝说明是否对针对其自身员工的威胁采取不同的处理方式。
这不仅是Anthropic的问题
Such issues are not limited to Anthropic. After a February school shooting in Tumbler Ridge, British Colombia, Canada, that killed eight people, OpenAI said it considered notifying Canadian police in June 2025, but concluded that the account did not show an imminent and credible risk of serious personal injury.
OpenAI stated in a policy post published in August 2025 that threats against others are routed to human reviewers, who may hand them over to law enforcement.
Heller's case is scheduled to start in November.
