Expired50% confidenceFactExact time
On platforms such as Meta and YouTube, AI models handle more than 90% of high-confidence cases automatically, routing boundary cases where the model is uncertain to human reviewers
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/9/2026
First Seen
Valid until: 10/7/2026(expired)
Sources
Why Can't AI Understand British Political Farce? The Cultural Comprehension Dilemma of Algorithms
hackernewshackernews7/8/2026
Related Claims
UnverifiedIn its 2022 transparency report, Meta disclosed that its AI systems automatically handle approximately 97% of violating content, with the remaining 3% handled by approximately 15,000 human reviewers.71% similarUnverifiedIn actual AI Agent project development, approximately 80% of the time is spent debugging prompts and troubleshooting abnormal Agent decisions, rather than on framework learning itself.67% similarUnverifiedAI can take over 80%–90% of repetitive software testing work, from test case design to coding to execution to result feedback.66% similarUnverifiedCompanies including Meta, Google, and Microsoft have successively disclosed internal AI coding penetration rate data, with some teams having more than 30% of their code generated by AI.66% similarUnverifiedAI Agents can handle over 80% of real-world business scenarios that cannot be pre-coded, due to their 'understanding-based' paradigm versus traditional 'rule-based' automation.66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/381006API
curl https://kongchang.com/api/v1/knowledge/claims/381006MCP
get_claim(id=381006)