Industry

Google’s AI Fails At 43% Of Finance Queries, Study Finds

2024-10-22 00:00:14

A study by The College Investor finds significant inaccuracies in Google’s AI-generated summaries for finance queries.

Out of 100 personal finance searches, 43% had misleading or incorrect information.

Key Findings

The study evaluated AI overviews across various financial topics, including banking, credit, investing, taxes, and student loans.

The results showed:

  • 57% of AI overviews were accurate
  • 43% contained misleading or inaccurate information
  • 12% were completely incorrect
  • 31% were either misleading or missing crucial details

Areas of Concern

Researchers noted that the AI struggled most with nuanced financial topics, such as taxes, investing, and student loans.

Some of the most concerning issues included:

  • Outdated information on student loan repayment plans
  • Incorrect details about IRA contribution limits
  • Misleading statements regarding 529 college savings plans
  • Inaccurate tax information that could potentially lead to penalties if followed

The AI handled basic financial concepts well but overlooked important exceptions and recent policy changes.

There are notable patterns in the queries Google’s AI got right versus those it got wrong.

Here are common themes.

Queries Google AI Got Right

  • Basic definitions and explanations: For example, “What is a wire transfer?” and “How does a credit card work?”
  • Simple, straightforward questions: Such as “Do I have to pay back student loans?”
  • Recent trending topics: Like “What was the Chase Glitch?”
  • General insurance questions: For instance, “When should I get life insurance?”

Queries Google AI Got Wrong

  • Complex tax topics: For example, “Can you use a 529 plan for a Roth IRA?” and “Does owning your house in an LLC help with taxes?”
  • Nuanced financial products: Such as “Is an IUL better than a 401k?”
  • Time-sensitive information: Like outdated student loan repayment plans or savings account rates.
  • State-specific financial rules: For instance, misrepresenting California’s 529 plan rules.
  • Questions requiring context-dependent answers: Such as “Can I file as independent for FAFSA?”
  • Queries about financial limits or thresholds: For example, incorrect IRA contribution limits.
  • Complex student loan topics: Particularly around forgiveness programs and repayment plans.
  • Investment comparisons: Like “Are annuities better than CDs?”

What This Means

Google’s AI performs well at giving straightforward answers to factual queries.

On the other hand, it struggles with nuanced understanding, up-to-date information, and consideration of multiple factors.

This suggests that the AI can handle basic financial literacy topics, but it’s unreliable for complex financial decisions or advice.

Potential Impact

Robert Farrington, founder of The College Investor, expressed concern about the findings, stating:

“If Google continues to present bad or misinformation about money topics to searchers, not only could it hurt their personal finances, but it could weaken already poor financial literacy in the United States.”

The study noted that following AI guidance could result in tax penalties or financial harm to consumers.

The College Investor believes Google should disable these AI-generated overviews for finance-related queries, especially those concerning taxes and investments.

Looking Ahead

Searchers must exercise caution when relying on AI-generated summaries for financial decisions.

When questioned about instances of misinformation, Google has previously stated, “the vast majority of AI Overviews provide high-quality information.”

The complete study, including detailed examples and methodology, is available on The College Investor’s website.


Featured Image: Koshiro K/Shutterstock