The proliferation of AI leaderboards incentivizes companies to optimize models for specific benchmarks. This creates a risk of "acing the SATs" where models excel on tests but don't necessarily make progress on solving real-world problems. This focus on gaming metrics could diverge from creating genuine user value.

Companies with valuable proprietary data should not license it away. A better strategy to guide foundation model development is to keep the data private but release public benchmarks and evaluations based on it. This incentivizes LLM providers to train their models on the specific tasks you care about, improving their performance for your product.

Public leaderboards like LM Arena are becoming unreliable proxies for model performance. Teams implicitly or explicitly "benchmark" by optimizing for specific test sets. The superior strategy is to focus on internal, proprietary evaluation metrics and use public benchmarks only as a final, confirmatory check, not as a primary development target.

The company actively works to prevent its answer engine from being gamed by "AI SEO" tactics. The core purpose is to maintain accuracy and trustworthiness; if a user can manipulate the results, that trust is broken. Perplexity views it as an arms race, stating they have "better engineers" to patch any hacks that so-called AI SEO firms might discover.

LM Arena, known for its public AI model rankings, generates revenue by selling custom, private evaluation services to the same AI companies it ranks. This data helps labs improve their models before public release, but raises concerns about a "pay-to-play" dynamic that could influence public leaderboard performance.

To maintain independence and trust, their public benchmarks are free and cannot be influenced by payments. The company generates revenue by selling detailed reports and insight subscriptions to enterprises, and by conducting private, custom benchmarking for AI companies, separating their public good from their commercial offerings.

Seemingly simple benchmarks yield wildly different results if not run under identical conditions. Third-party evaluators must run tests themselves because labs often use optimized prompts to inflate scores. Even then, challenges like parsing inconsistent answer formats make truly fair comparison a significant technical hurdle.

Don't trust academic benchmarks. Labs often "hill climb" or game them for marketing purposes, which doesn't translate to real-world capability. Furthermore, many of these benchmarks contain incorrect answers and messy data, making them an unreliable measure of true AI advancement.

To maintain trust, Arena's public leaderboard is treated as a "charity." Model providers cannot pay to be listed, influence their scores, or be removed. This commitment to unbiased evaluation is a core principle that differentiates them from pay-to-play analyst firms.