Deepfakes, social engineering push identity checks into live conversations

AI is a powerful tool for fraud on its own, but social engineering sharpens its edge. People remain easy to fool. As a recent article in Security Week puts it , “human defenders should not and cannot be expected to detect trickery designed to manipulate their psychology. And the tricks are becoming better hidden and more sophisticated with the use of AI deep fakery.”

As agents flood digital ecosystems, it’s not going to get any easier to verify who’s who, who’s doing what and whether or not a request is valid. The identity sector is responding by looking past one-time checks to a more comprehensive monitoring solution that uses biometrics in concert with other tools, to meet the demands that AI fraud and agentic misadventure have introduced.

Netarx harnesses thousands of signals to protect vast attack surface

Michigan-headquartered Netarx is among the host of firms exploring how to address the social engineering problem with technology. The company has announced the launch of its new AI Defense API, which a release calls a “comprehensive platform that detects and neutralizes deepfakes, social engineering, identity spoofing, and fraud across every channel of corporate communication.”

The tool is built around the Netarx AI Defense Meta Harness orchestration layer, which uses 40 specialized AI models to analyze digital signals in real time, cross-validating thousands of signals across modalities to catch sophisticated, multi-channel attacks.

Sandy Kronenberg, CEO and Founder of Netarx, says that “for decades we protected the network. Now the attack surface is the conversation itself.”

Elements of the conversation, however, can be analyzed. For hardware, device fingerprints, EXIF data, compression signatures, and location mismatches; for audio, micro-artifacts undetectable by the human ear. For faces, it “inspects individual video frames for anomalies in micro-expressions, unnatural blinking frequencies, lighting inconsistencies, or warping around hair lines.”

Kronenberg says “AI Defense API fundamentally changes the equation. We’re not just protecting human-to-human interactions from deepfakes and social engineering , we’re defending the emerging world of non-human-to-human communication, where AI agents and automated systems interact with your people every day.”

Onecom launches new version of Vox voice deepfake, IDV tool

Onecom Technology Group has released the generation of its Vox voice platform , which combines deepfake call detection, identity verification, recording and transcription with actionable intelligence gathered from call records, according to a release.

The rebuild adds technology that can detect synthetic voices in real-time, and automatically identify known callers from their unique voice fingerprint.

Aaron Brown, CEO of Onecom Technology Group, says the company “wanted to give businesses practical tools to detect threats while also helping them get more value from every customer conversation through a system that learns from calls, providing businesses with intelligence they can act upon to improve productivity.”

Per the release, the new version of Vox is available now to all Onecom customers and will be launched to Onecom Partners resellers in Q4 2026.

Pindrop Botstopper identifies machine callers using AI Voice Consortium

Pindrop has launched its Pindrop BotStopper technology, a standalone offering that, according to a release, detects AI agents and automated callers in real time.

Pindrop is responding to shifts in the market with a broader expansion of its Identity Trust Platform across customers, workforces and AI agents . Company CEO Vijay Balasubramanian says that, with BotStopper, “the goal is not to block AI agents. Legitimate agents are going to call enterprises on behalf of real customers and they should be welcomed.”

“The goal is to know which agent is calling, who it represents and what it should be allowed to do. Pindrop BotStopper is the first step. It tells you in two seconds that the caller is a machine and the AI Voice Consortium tells you which one. Everything we build from here answers the questions that follow.”

The new standalone tool is powered by the AI Voice Consortium, “a new registry of more than 5,000 AI voices in a Pindrop-managed database,” which “brings ‘know your agent’ to the voice channel, the same way payment networks and identity providers are now verifying AI agents in digital transactions.”

The company says results are already showing up in production: “at one leading healthcare organization, bot calls peaked at 6,668 in a single month, when approximately $15.1 million in HSA balances were exposed to risk. Within four months of deploying Pindrop’s real-time bot detection, that activity had fallen by 94.3 percent.”

Pindrop is also expanding its identity trust capabilities deeper into the enterprise, adding continuous video analysis during screen sharing to Pindrop Pulse for Meetings , and showcasing new agentic capabilities including a Fraud Investigation Agent and an Enrollment Agent.

AI fraud | continuous verification | deepfake detection | deepfakes | Netarx | Onecom | Ping Identity | voice biometrics