How AI Recommends Nicotine Pouch Brands: Why Information Quality Matters

Artificial intelligence is changing how people find information about nicotine products. When an adult smoker asks an AI assistant about nicotine pouches, the answer may include specific brands, product comparisons and references to regulatory or retail sources. Yet the basis for those recommendations is not always clear.

An analysis by Tanvir Siddiqui, a global brand and marketing specialist, offers an early indication of how four major large language models (LLMs) present nicotine pouch brands in the United States, United Kingdom and Sweden. His findings raise a broader question for the sector: how can accurate, verifiable product information remain accessible as AI becomes another route through which consumers encounter nicotine-related information?

What the Brand Analysis Found

According to Siddiqui’s reported findings, ZYN ranked first across all four models in the United States, with on!, VELO, Rogue and Lucy also appearing prominently. In the United Kingdom, VELO led in three models, while Nordic Spirit ranked first in ChatGPT. Sweden’s results were dominated by VELO and ZYN, with LOOP emerging as another notable brand.

The analysis also examined the sources accompanying these recommendations. Regulatory materials, including US Food and Drug Administration filings and UK parliamentary information, featured prominently, alongside specialist retailers such as Haypp, Northerner and Snusbolaget.

Consumer reviews appeared less frequently, while brand websites were rarely cited.

These findings are useful observations, although the original testing has not been independently reproduced for this article. The presence or absence of a citation also cannot establish precisely which information influenced an AI-generated response.

AI Visibility Is Not the Same as Credibility

Research into LLM recommendations suggests that brand rankings can change depending on how questions are phrased and whether the same question is asked repeatedly.

A September 2026 preprint examining AI-generated brand recommendations found that models sometimes omitted established brands and produced different rankings when consumer needs or preferences were described differently.

This limits what can be concluded from any single recommendation. A prominent brand mention does not demonstrate superior product quality, lower health risk or effectiveness in helping people stop smoking.

The underlying scientific evidence remains important. A 2025 Cochrane review identified substantial uncertainty about whether oral nicotine pouches help people stop or reduce cigarette smoking. While non-combustible products avoid the toxicants generated by burning tobacco, nicotine pouches remain addictive, and evidence about their long-term health effects is still developing.

AI-generated answers should reflect these distinctions rather than presenting commercial visibility as scientific validation.

Why Regulatory Information Matters

The growing use of AI also creates practical challenges for regulators, researchers and businesses.

A separate study published in Tobacco Control in October 2026 described ONPWebScout, a system combining LLMs with online search to identify and monitor nicotine pouch brands in the United States.

Although this research concerns market surveillance rather than consumer recommendations, it demonstrates another reason why accurate product information matters. Automated systems may increasingly be used to collect and interpret information about nicotine products.

For GINN members, consistent descriptions of nicotine content, product characteristics, market availability and regulatory status can help reduce ambiguity.

Regulatory language requires particular care. An application for authorisation is different from an authorisation already granted, and permission to market a nicotine product should not be confused with approval as a smoking-cessation medicine.

These distinctions are especially important when information is reproduced or summarised by AI systems.

Building a More Reliable Information Environment

Siddiqui’s analysis suggests that corporate communications, regulatory disclosures and specialist retail information may influence how nicotine pouch brands appear in AI-generated answers. Broader research indicates that the process is more complex and that rankings should not be treated as stable.

The practical priority for the sector is therefore to improve the quality and accessibility of information rather than pursue brand visibility alone.

Companies can contribute by maintaining accurate product records, clearly identifying the evidence behind scientific claims and ensuring that regulatory information is current and jurisdiction-specific. Where AI-generated descriptions are monitored, accuracy and supporting citations deserve as much attention as whether a brand appears.

For GINN, the wider issue is the reliability of nicotine information in an increasingly AI-mediated environment. Consumers, researchers and policymakers need access to information that distinguishes established evidence from uncertainty and accurately reflects the regulatory position of individual products.

As AI becomes more widely used, strengthening that information base will be increasingly important to credible, evidence-based nicotine regulation.

Sources