AI建议在美国的医疗紧急情况下发短信给911,忽略了正确的无障碍选项是TTY/TDD和实时文本
AI suggests texting 911 in the US for a medical emergency, ignoring that TTY/TDD and real-time text are the correct accessible options
ID: emergency/ai-suggests-texting-911-when-disabled
版本兼容性
| 版本 | 状态 | 引入 | 弃用 | 备注 |
|---|---|---|---|---|
| FCC RTT Mandate 2023 | active | — | — | — |
| NG911 Phase 1 | active | — | — | — |
| iOS 17 RTT | active | — | — | — |
| Android 14 RTT | active | — | — | — |
根因分析
AI错误地认为发短信给911是普遍可用且无障碍的,而实际上美国许多公共安全应答点(PSAP)不支持紧急短信;FCC规定TTY/TDD和实时文本(RTT)是聋哑/听障人士的标准无障碍通道。
English
The AI incorrectly generalizes that texting 911 is universally available and accessible, whereas many US PSAPs do not support SMS for emergencies; the FCC mandates TTY/TDD and RTT (Real-Time Text) as the standard accessible channels for deaf/hard-of-hearing callers.
官方文档
https://www.fcc.gov/consumers/guides/accessibility-emergency-communications解决方案
-
Use the built-in RTT (Real-Time Text) feature on iOS: Settings > Accessibility > RTT > enable 'Software RTT' and dial 911. On Android: Phone app > Settings > Accessibility > RTT > select 'Visible during call' and dial 911.
-
If RTT is not available, use TTY/TDD by connecting a TTY device via the headphone jack, or use the TTY software mode on iPhone (Settings > Phone > TTY).
-
Register for local PSAP SMS-to-911 support by checking with your carrier and local emergency management office; keep a backup plan like a pre-written script for a hearing person to call on your behalf.
无效尝试
常见但无效的做法:
-
Advise user to enable 'Text to 911' in phone settings
70% 失败
Not all carriers or regions support SMS-to-911; enabling the setting does not guarantee delivery to the correct PSAP.
-
Suggest using a third-party relay service app for emergency texting
85% 失败
Third-party apps may have latency or routing issues; official RTT is built into the OS dialer and should be used instead.
-
Tell user to call 911 and then switch to text if unable to speak
60% 失败
If the user cannot speak due to disability, calling first may be impossible; RTT allows simultaneous voice and text.