When Grok Mistakenly Sends $200,000 in Crypto: Lessons for Enterprise AI Visibility

A $200,000 Incident That Reveals the Flaws of Generative AI

In May 2025, an X user successfully manipulated Grok, xAI's AI system, into sending $200,000 in cryptocurrency. His method: Morse code hidden within his queries to bypass security filters. The incident went viral across social media with over 2,000 reactions on Reddit within hours.

For business leaders, this event goes beyond a mere technical anecdote. It illustrates a reality that many underestimate: generative AI systems like Grok, ChatGPT, and Perplexity have become major interfaces between brands and their potential customers. And these interfaces can be manipulated, misinformed, or simply incorrect about your company.

This article analyzes what the Grok/crypto case teaches us about the risks and opportunities of AI visibility for SMEs and mid-market companies. You'll discover how these systems work, why your company may be misrepresented, and what concrete actions to implement to appear correctly in Grok and other generative engines.

How Grok and Generative AI Systems Build Their Responses

Direct Access to X Data and Web Content

Grok distinguishes itself from other AI systems through its native integration with X (formerly Twitter). Unlike ChatGPT, which primarily relies on training data and occasional web searches, Grok accesses real-time posts, trends, and conversations from Elon Musk's platform.

This architecture has direct consequences for businesses. If your brand is mentioned on X, Grok can integrate these mentions into its responses, whether positive or negative. A viral thread criticizing your customer service, a sarcastic comment from an influencer: everything becomes raw material for the AI.

The Source Verification Problem

The $200,000 crypto incident reveals a fundamental flaw: Grok, like its competitors, doesn't systematically verify the legitimacy of instructions it receives. The user exploited this weakness by encoding his commands in Morse code, thus bypassing textual security filters.

For your business, the risk is similar but reversed. Grok can propagate incorrect information about you simply because it's circulating on X or the web. An outdated price never updated, an obsolete address, an abandoned product presented as available: these errors spread with each user query.

Concrete Risks of AI Misinformation for SMEs and Mid-Market Companies

Three Scenarios We Regularly Observe

At AISOS, our AI visibility audits reveal recurring problems among French and Belgian B2B companies:

• Outdated information: 67% of audited companies have at least one critical piece of incorrect data in ChatGPT or Perplexity responses (revenue, employee count, service offerings)
• Incorrect attribution: the AI attributes characteristics of competitors with similar names or companies in the same sector to your business
• Complete absence: for relevant industry queries, your company doesn't appear while less established competitors are cited

Impact on the B2B Buying Journey

B2B decision-makers increasingly use generative AI in their research process. According to a 2024 Gartner study, 41% of B2B buyers consult ChatGPT or similar tools before contacting a potential supplier. For the 30-45 age group, this figure rises to 58%.

An incorrect AI response about your company can therefore eliminate your candidacy before you're even aware of the opportunity. The prospect simply won't contact you, convinced by the AI that you don't match their needs.

Why Grok Represents a Specific Challenge for Businesses

The Amplifying Effect of X on Reputation

Grok mechanically amplifies what's said about you on X. If your company is active there and generates positive conversations, it's an asset. If you're absent or mentions are negative, Grok will reflect this reality in its responses.

This mechanism creates a feedback loop: users who query Grok about your sector receive responses influenced by X, then potentially share these responses on X, reinforcing the cycle.

The Question of Moderation and Bias

The crypto incident also illustrates Grok's moderation limits. Elon Musk has deliberately positioned his AI as less censored than its competitors. This approach has advantages in terms of free speech but creates risks for businesses.

Grok can relay criticism, rumors, or unverified information with fewer filters than ChatGPT or Claude. For an SME unfairly targeted by disparagement on X, this characteristic poses a direct reputational problem.

How to Appear Correctly in Grok: Practical Recommendations

Step 1: Audit Your Current Presence

Before any action, measure the actual state of your AI visibility. Query Grok, ChatGPT, Perplexity, and Gemini with typical queries:

• "What does [your company name] do?"
• "Who are the best [your profession] in [your region]?"
• "[Your company name] customer reviews"
• "Difference between [your company] and [competitor]"

Document the responses, identify errors and gaps. This initial mapping will guide your priorities.

Step 2: Structure Your Data for LLMs

Generative AI systems extract their information from structured sources. To improve your visibility, focus on:

• Your website: clear pages with direct statements about your activity, clients, location. Avoid empty marketing jargon
• Structured data: schema.org Organization, LocalBusiness, tags that help AI understand your factual information
• Wikipedia and authoritative sources: if your company is eligible, a properly sourced Wikipedia page strongly influences AI responses
• LinkedIn: complete company page with precise description, often used as a verification source by LLMs

Step 3: Activate Your X Presence for Grok

Since Grok feeds directly from X, your strategy on this platform impacts your AI responses. Priority actions:

• Verified company account with optimized description
• Regular posts about your industry expertise
• Engagement with conversations in your sector
• Monitoring mentions to respond to incorrect information

The goal isn't to go viral, but to create a corpus of content that Grok can use to describe you correctly.

Step 4: Monitor and Correct Continuously

AI visibility isn't a one-time project. Models evolve, their sources change, your competitors optimize their presence. Set up monthly monitoring of your critical queries and document the evolution of responses.

What the Grok Incident Teaches Us About the Future of AI Visibility

AI Systems Are Becoming Reputation Gatekeepers

The user who obtained $200,000 exploited a technical flaw. But the incident mainly reveals the growing power of AI as trusted intermediaries. Users trust Grok to execute financial actions: they'll equally trust it to recommend a B2B supplier.

This evolution places AI visibility at the same strategic level as SEO fifteen years ago. Companies that ignore it will gradually lose opportunities to competitors better positioned in generative responses.

Regulation Will Transform Practices

The European AI Act, which entered progressive application since 2024, imposes transparency obligations on AI providers. Eventually, companies could obtain rights to correct incorrect information about them. Meanwhile, the best protection remains proactively providing AI systems with correct and verifiable information.

Grok, ChatGPT, Perplexity: A Multi-Platform Strategy Is Necessary

Each generative AI has its preferred sources. Grok favors X, Perplexity abundantly cites recent web sources, ChatGPT relies on its training and plugins. An effective AI visibility strategy covers all these channels, without favoring a single engine.

Summary: Transforming Risk into Competitive Advantage

The $200,000 crypto incident isn't just a technical curiosity. It illustrates the power and flaws of generative AI systems like Grok—systems that already influence how your prospects perceive your company.

SME and mid-market company leaders must integrate three realities:

• Your customers already use these tools to research you, whether you want it or not
• Information can be incorrect and you won't be informed
• Optimization is possible with concrete actions on your content and digital presence

AI visibility is no longer a forward-looking topic. It's an immediate operational challenge for any company that depends on its reputation and ability to be found by potential customers.

To evaluate your current situation and identify your action priorities, AISOS offers AI visibility audits tailored to French-speaking SMEs and mid-market companies. Contact us to discover how your company actually appears in responses from ChatGPT, Grok, Perplexity, and Gemini.