Originally published on The Searchless Journal
Americans lost $893 million to AI-powered scams in 2025. Globally, fraud losses hit $400 billion. And the fastest-growing weapon in the scammer's arsenal is voice cloning, where AI generates such convincing replicas of someone's voice that even close family members cannot tell the difference.
Google's response launched this week. Fake call detection, rolling out now on Android, uses an RCS-based digital handshake to verify whether an incoming call is actually coming from the person it claims to be. If the caller's device cannot cryptographically confirm its identity, your phone displays a warning.
It is a consumer protection feature. It is also something more.
The verification infrastructure Google is building to fight AI fraud is the same pattern that will determine which sources AI engines trust, which brands they recommend, and which businesses disappear from AI-generated answers entirely. Cryptographic trust signals. Device-level authentication. AI systems verifying other AI systems in real time.
If Google can verify a phone call is real, it can verify your brand is real.
Here is why that matters for every business trying to be visible in AI search.
The Scale of the Problem
AI voice cloning has moved from a novelty to a weapon. The FBI's Internet Crime Complaint Center reported $893 million in losses from AI-enabled scams in 2025, a figure that has more than doubled year over year. INTERPOL's March 2026 Global Financial Fraud Threat Assessment put total global fraud losses at $400 billion, with AI-generated impersonation scams cited as the fastest-growing category.
CNN reported in May that AI audio deepfakes have become so realistic that most people cannot distinguish them from genuine recordings in controlled tests. A 30-second voice sample is enough to clone someone's voice with high fidelity using commercially available tools.
The scams follow a predictable pattern. A grandparent receives a call from someone who sounds exactly like their grandchild, claiming to be in an emergency and needing money wired immediately. An executive gets a voicemail from what sounds like their CEO, instructing an urgent wire transfer. A employee receives a call from their IT department asking them to reset credentials on a phishing page.
The voice is real enough to fool anyone who is not specifically trained to detect it.
How Google's Fake Call Detection Works
Google's solution is elegantly simple at the user level and technically sophisticated underneath.
When a call comes in, the Phone by Google app checks whether the caller's device can send a silent RCS-based confirmation signal. This signal is encrypted end-to-end using the RCS messaging standard, the same protocol that powers modern Android messaging. If the caller's device sends the confirmation and it matches, the call is verified. If the signal is missing or does not match, your phone displays a warning.
Both parties need to use the Phone by Google app for verification to work. The feature is built on an open standard, meaning other messaging and calling apps can adopt the same verification protocol.
The key insight: this is not pattern matching or AI detecting anomalies in the audio stream. It is a cryptographic handshake at the device level. The verification does not rely on analyzing whether the voice "sounds right." It relies on confirming that the device sending the call is the device associated with the claimed identity.
This is a fundamentally different approach to trust. Instead of analyzing content (does this voice sound real?), it verifies provenance (did this device actually send this call?).
The Parallel: From Call Verification to Source Verification
Here is where the consumer protection story connects to the AI visibility story.
Google's call verification works by establishing a chain of trust: device identity, encrypted signal transmission, matching against known identity records. The call content is secondary. What matters is whether the source is authentic.
Now apply that same pattern to web content and AI citations.
When Google's Gemini generates an AI Overview or an AI answer, it synthesizes information from multiple web sources. The question of which sources to trust is central to the quality of that answer. Google has traditionally used signals like PageRank, E-E-A-T (Experience, Expertise, Authoritativeness, Trustworthiness), and helpful content signals to rank sources.
But AI answers require a different level of trust verification. An AI Overview typically synthesizes 3 to 8 sources into a single answer. Each cited source needs to be authoritative enough to anchor a claim that Google's AI is making directly to the user. The bar is higher than traditional search ranking because Google is not just linking to the source. It is endorsing the source's claims by including them in its own generated answer.
The verification infrastructure Google is building for call authentication maps directly onto the verification infrastructure needed for source authentication:
Device identity maps to domain identity. Just as Google verifies that a calling device matches a known identity, it can verify that a domain matches a known entity in its Knowledge Graph. Consistent entity data, verified business listings, and structured authorship information become the domain-level equivalent of device authentication.
Encrypted signal transmission maps to structured trust signals. The RCS handshake is a machine-readable signal that confirms authenticity. Schema.org markup, JSON-LD structured data, and verified author profiles are the web equivalent: machine-readable signals that confirm a source's authority and expertise.
Content analysis maps to provenance verification. Google's call verification does not analyze whether the voice sounds real. It verifies whether the device is real. Similarly, AI citation is moving beyond analyzing whether content "looks authoritative" toward verifying whether the source has a proven track record of expertise in its domain.
What This Means for Brand Visibility
The convergence of fraud verification infrastructure and AI citation methodology creates a clear imperative for brands. The AI engines that will dominate discovery in the next few years are building trust verification into their core architecture. Brands that pass trust verification will be cited. Brands that do not will be invisible.
Here are the trust signals that matter right now:
Consistent entity data across the web. Your brand's name, address, phone number, and business category should be identical across Google Business Profile, social media, industry directories, and your own website. Inconsistencies are the equivalent of a failed cryptographic handshake: they signal that something is not quite right about this entity.
Verified authorship and expertise signals. Google's E-E-A-T framework has always emphasized author expertise. In the AI citation era, this becomes even more important. Content attributed to named, credentialed authors with verifiable expertise in their domain is more likely to be cited than anonymous content or content from unverified sources.
Structured data implementation. Schema.org markup and JSON-LD structured data are the machine-readable trust signals that AI engines use to understand what your content is about, who created it, and why it is authoritative. Think of structured data as the RCS handshake of web content: a machine-readable signal that confirms your content's identity and provenance.
Authoritative backlink profiles. The traditional signal of a source's authority remains important. But the quality bar is higher for AI citation than for traditional ranking. A few links from genuinely authoritative sources (industry publications, academic institutions, government websites) carry more weight than thousands of low-quality directory links.
Freshness and consistency. Google's fraud detection systems flag anomalies. Similarly, AI citation systems flag sources that are inconsistent or outdated. Regularly updated content that demonstrates ongoing expertise in a domain signals that a source is active and current.
Cross-platform presence. Google's verification infrastructure extends across Search, Android, Drive, Workspace, and its AI features. Brands that have a consistent, verified presence across multiple Google surfaces build a stronger trust profile than brands that only exist on one surface.
The Bigger Picture: AI-to-AI Verification
Google's fake call detection is an early example of what will become a pervasive pattern: AI systems verifying other AI systems.
On the consumer side, Google's AI verifies that an incoming call is not an AI-generated deepfake. On the enterprise side, Google's AI verifies that a web source is not AI-generated spam or unreliable content. The verification methodology is the same: establish identity, confirm provenance, verify authenticity through cryptographic or structured signals.
This has profound implications for the AI visibility industry. The current approach to GEO (Generative Engine Optimization) focuses heavily on content optimization: writing comprehensive articles, using natural language, covering topics thoroughly. These tactics work today. But as AI verification infrastructure matures, the content itself will matter less than the trust signals surrounding the content.
A mediocre article on a highly trusted, well-verified domain will outperform a brilliant article on an unverified domain. Not because the content is better, but because the trust infrastructure confirms that the source is real, authoritative, and worth citing.
Brands that invest in trust signals now are building the foundation for long-term AI visibility. Brands that focus only on content optimization are building on sand.
The Numbers Behind the Urgency
The fraud data tells the story of how fast AI-generated content can erode trust. If $893 million was lost to AI scams in 2025, the volume of AI-generated deceptive content on the web is orders of magnitude larger. Google is not building verification infrastructure as an academic exercise. It is building it because the alternative is an information ecosystem where no source can be trusted.
For brands, the urgency is real. Google's AI Overviews now appear on more than 80% of B2B queries. Gemini is embedded in Android, Workspace, Drive, and Google Search. The verification infrastructure being tested in call detection will be applied to web sources. The question is not whether this will happen, but how quickly.
Brands that establish strong trust signals now will be positioned favorably when the verification layer activates for web content. Brands that wait will be playing catch-up in an environment where trust verification is automated and instantaneous.
What to Do Right Now
If you are responsible for your brand's visibility in AI search, here is a practical checklist aligned with the emerging trust infrastructure:
Audit your entity consistency. Search for your brand across Google, Bing, social media, and industry directories. Is your business information identical everywhere? If not, fix it. Every inconsistency is a failed trust signal.
Implement comprehensive structured data. If you are not using schema.org markup and JSON-LD structured data on every key page, you are missing the machine-readable trust signals that AI engines use to verify your content. This is the web equivalent of the RCS handshake.
Establish verified authorship. Every piece of content your brand publishes should be attributed to a named author with verifiable credentials. Author pages with biographies, credentials, and links to published work build the expertise signals that AI citation systems rely on.
Build authoritative backlinks, not volume. A single link from a respected industry publication is worth more than a hundred directory submissions. Focus on earning citations from sources that AI engines already trust.
Keep your content fresh and consistent. AI verification systems, like fraud detection systems, flag anomalies. Content that is regularly updated and consistently high-quality signals an active, authoritative source.
Measure your AI visibility. Use an AI visibility audit to see how your brand appears across AI answer engines like Google AI Overviews, Perplexity, and ChatGPT. You cannot optimize what you do not measure. Check your AI visibility here.
The Trust Infrastructure Is Being Built Now
Google's fake call detection is a consumer feature that reveals an enterprise truth. The infrastructure for verifying AI-generated content is being built right now, starting with phone calls and expanding to every surface where AI interacts with information.
The $893 million lost to AI scams in 2025 is the consumer-facing symptom. The enterprise-facing reality is that trust verification is becoming the gating factor for AI visibility. Brands that understand this and invest in trust signals will be the ones that AI engines recommend. Brands that ignore it will be the ones that AI engines cannot verify, and therefore cannot trust, and therefore will not cite.
The handshake is happening. The question is whether your brand is sending the right signal.
Top comments (0)