Asif Waliuddin is the Founder and CEO of NXTG.ai, where he leads strategic initiatives in enterprise AI transformation.
Sleeper poisoning planted fabricated memories in up to 99.8% of attempts on GPT-5.5, so AI memory must prove it is current before it is obeyed.
KPMG pulled an AI report when only 5 of its 45 citations held up. The deeper failure is unverified AI output becoming trusted organizational memory.
Part 2 of my Jev test, pre-registered before the first call: up to five similar past records added to every prompt. Opus went from 23% to 73% on one routing question, and neither model showed a clear effect on another. Per class results for Jev, Claude Opus 5.5 and memory alone.
A pre-registered test of TypeSafe's Jev against Claude Opus 5.5 and a local Qwen3-14B on 340 decisions with known answers: accuracy, calibration, cost and speed, per class.
Faultline started as a four-day Kaggle entry with zero tests. Faultline Pro now has 4,887 counted tests. These are the failures that put them there.
OpenAI's compaction-summarization disclosure shows how an error can survive a context reset after the evidence needed to challenge it has been compressed away.
