Which companies are faking AI?+
Check the filing, not the press release. A company that says AI 200 times in its 10-K while R&D spending falls is marketing, not investing. GPTTicker counts AI mentions per 10-K for 499 S&P 500 companies and pairs each count with an AI-wash score, so the talkers stand out fast. Palo Alto Networks led the latest filings with 244 mentions; 3M had 16. Same market, very different stories.
What is an AI washing ETF, and how do I avoid one?+
An AI washing ETF sells artificial intelligence exposure but holds companies with thin AI substance. The fix is looking through the fund at the holdings. TrueAI scores any AI ETF's top 10 holdings free, using 10-K mention counts and R&D. If half the top 10 barely mention AI in their filings, you are paying AI prices for a regular tech fund.
QQQ vs AIQ vs BOTZ: which has the most real AI exposure?+
QQQ is a Nasdaq-100 fund wearing no AI costume, which is honest. AIQ and BOTZ charge you for AI exposure, so they should be judged on it. Run each fund's holdings through the AI-wash scores: count how many top holdings show rising AI mentions plus real R&D. The fund with more substance per dollar of expense ratio wins. Names lie; filings do not.
Is there a tool that tracks AI mentions in earnings calls?+
A few. Quartr has the transcripts, and our database tracks 10-K mentions which are harder to spin than call remarks. The honest answer: earnings calls are theater, 10-Ks are sworn statements. Track both, but weight the filing.
How should I compare AI ETF expense ratios?+
Divide substance by cost. A 0.68% AI ETF full of AI washers is worse than a 0.20% Nasdaq fund with real AI names inside. Stack the fee against the AI-substance score: cheap hype is still hype.