Chatcontrol 2025 w skrócie, po polsku - Bert Hubert's writings
a day ago
- EU member states, led by Denmark, want to force WhatsApp, Signal, and similar apps to scan all photos and links using AI for child pornography.
- If AI has any doubt about content, the photo, location, phone number, and other data are reported to Europol and local police, with potential for false positives and language barriers.
- WhatsApp/Meta would create the AI to monitor users on behalf of the EU, with a low threshold for reporting (only if 'obviously unfounded' can be avoided).
- Even if authorities deem a photo 'legal', a copy is retained in a database for possible future use.
- Users will be asked to consent to AI scanning; refusal means losing the ability to share photos and links. Signal has threatened to leave the EU if forced to comply.
- Previous versions of the proposal did not include AI; this element was reintroduced recently.
- Police and child protection experts are already overloaded, yet politicians focus on surveillance rather than providing more resources, training, or international cooperation.
- The proposal creates a massive surveillance center run by Americans, and history shows such mechanisms are eventually used for other purposes, while exempting police, military, and national security.
- The German government appears likely to support the proposal, along with other governments, prompting calls for citizens to contact politicians and raise awareness.