Telegram Policy Violations?

Following up on the previous issue of generic ad rejections on Telegram, we're now sometimes getting slightly more specific feedback, but it's still unactionable. The primary culprit now seems to be 'Telegram policy violations' without further detail, making it impossible to diagnose. We've meticulously reviewed the official Telegram Ad Platform guidelines and terms of service, cross-referencing them with best practices and common rejection patterns from other major ad platforms like Facebook and Google Ads. We’ve experimented with a wide array of ad creatives – from ultra-minimalist text-only ads to those incorporating emojis and varied calls-to-action, alongside testing different landing page structures (simple one-pagers vs. detailed product pages, even different domains). We’ve also employed AI-driven content analysis tools to pre-screen ad copy for any potentially sensitive keywords, aggressive language, or implied claims that might trigger automated flags. All landing pages are strictly HTTPS, optimized for rapid loading, and are fully mobile-responsive, with double and triple checks for any hidden redirects or cloaking attempts. Despite these efforts, the rejections continue with phrases like 'Violation of Telegram Ad Platform policies,' 'Content deemed inappropriate,' or 'Misleading claims detected' – even when the content is factually accurate and benign. Our current working hypotheses delve into the underlying algorithmic mechanics. Is there an undocumented, heuristic algorithm specifically targeting 'Telegram policy violations' that flags subtle linguistic patterns, specific character combinations, or even sentence structures that aren't explicitly forbidden? Could account history or IP reputation play a role, even for seemingly new and clean accounts? We're also considering if certain keyword combinations, even innocuous ones, might trigger an automatic cascade of rejections. Furthermore, the possibility of inconsistent human review, where context is misinterpreted or subjective biases influence decisions, remains a concern, especially concerning the rigor of Telegram's ad policy enforcement. We're seeking insights into any deeper, non-obvious causes for these persistent 'Telegram policy violations' and practical strategies to debug this effectively. Are there any known undocumented pitfalls, specific content nuances, or common misinterpretations of policy that the community has encountered? Has anyone faced this level of ambiguity with Telegram policy violations and found a breakthrough?