50-company study: median enterprise knowledge base doc is 505 days stale, 1 in 65 AI citations hits expired verification
femke_plantinga · x · 2026-11-18
A study of 50 company knowledge bases (14,300 people), 16 team interviews, and custom AI search evals reveals how poorly enterprise knowledge serves AI agents:
- Stale content: the median knowledge base holds 5,175 docs, half older than 505 days; docs are rarely edited after publishing, many go a year with zero readers.
- Head-heavy: a tenth of people write most of it; the top 10% of pages (500) get most reads.
- Agents are now readers: agents make 1 in 5 document reads; 23% of members query team knowledge via outside AI tools.
- Verification is broken: only 1.3% of docs carry a current verification; 1 in 65 AI citations lands on an expired one.
- Contradictions everywhere: 57% of 300+ audited public help centers contradicted themselves at least once, and agents pick one answer with no warning.
Takeaway: most enterprise knowledge feeding agents is stale, unverified, and self-contradictory — an underapprecized quality risk for AI search and agent deployments.
More from coding & agent
- Using Google Antigravity agent swarms to analyze the November 2026 Core Update — gaganghotra_ · 2026-11-18
- Dev builds MCP skills tool so Claude, Cursor and Copilot share one set of instructions — EntrepreneurOk5764 · 2026-11-18
- 12 indie AI mini-game devs share real revenue: top Steam title made $42K RMB in 49 days — sujingshen · 2026-11-18
- Oracle agent memory cuts input tokens ~10x at turn 80 vs full chat replay — bibryam · 2026-11-18
- Huawei researcher's Cambridge talk: agent-era code bottlenecks shift to specs and verification — satnam6502 · 2026-11-18
- Plane AI now writes PQL queries for you, keeping them editable — JosephJacks_ · 2026-11-18