We need to talk about filters: algorithmic copyright enforcement vs data protection.
summary
Article 17 makes platforms primarily liable for user uploads, pushing them toward upload filters and fingerprinting systems like Content ID or third‑party matchers (Audible Magic), which reliably match exact files but fail to assess context or legal exceptions, producing high false positives and freedom‑of‑expression risk; implementation triggers extensive personal data processing and GDPR questions, stakeholder dialogues stall, some states draft divergent national rules, and alternatives discussed include collective licensing and expanded human review—current tools work for exact matches but not for nuanced legal judgments.
AI-written summary · about
chapters (8)
- 0:00Panel introduction
- 3:24Article 17 overview
- 10:04Data protection and privacy
- 15:43Implementation challenges
- 22:06Platform liability debate
- 27:30Copyright and platform power
- 36:42Audience Q&A
- 59:06Technical realities and wrap
related talks
01h 09mStrip-searching the internet - upload filters and generalised monitoring ...
16mReflections on recent privacy trends
01h 14m'We give up -please fix the interwebz'. ...
01h 11mTransforming GDPR into a Risk Based Harm Tool Alongside Specific AI Regulation...
01h 15mTHE ROLE OF TRANSPARENCY, USER CONTROL, AND THE BALANCE BETWEEN FUNDAMENTAL RIGHTS
01h 15mCALIBRATING THE AI ACT - IS IT THE RIGHT FRAMING TO PROTECT PERSONAL AND FUNDAMENTAL RIGHT
01h 14mNavigating regulatory complexities Enforcing data protection in the field of AI
01h 18mOnline privacy versus freedom of speech. Balancing rights in the EU context.