Observed Signal · Sep 28, 2026 · Product Launch · Source: Tech.eu (European Tech & Deals) · Impact: 2/5 · Sentiment: Positive
Interhuman AI teaches AI to read non-verbal human signals
Copenhagen-based Interhuman AI has launched Inter-2, a model that detects 12 social signals such as hesitation, uncertainty, and agreement in real-time from text, audio, and video. The startup aims to introduce Artificial Social Intelligence, enabling AI to understand non-verbal cues. Co-founded by CEO Paula Petcu, COO Frederik Sally, and CRO Line Clemmensen, the company is building its own models in Europe. Inter-2 delivers up to four times faster inference with improved benchmark performance. The technology is targeted at applications in corporate training, digital health, sales coaching, and market research. Interhuman emphasizes transparency and traceability of model outputs, adhering to GDPR and EU AI Act. The company is expanding into voice-only interactions and will announce two additional models in October.
This is a niche AI model launch focused on non-verbal communication analysis, which could impact AdTech by enabling more nuanced consumer interaction analysis, but it is not a major industry shift.
Track Antler Signals & Market Shifts in Real-Time
Polaris7 autonomous intelligence agents track regulatory filings, primary sources, executive changes, and deal flow 24/7. Create your Polaris7 workspace to monitor these entities.
Key Takeaways & Evidence Grounding
- Interhuman AI launched Inter-2, a model detecting 12 social signals in real-time from text, audio, and video.
- Inter-2 delivers up to four times faster inference with improved benchmark performance.
- The company is developing its own AI models in Europe and is GDPR-compliant, undergoing security certification audits.
- Interhuman is working with digital health clients, including the Centre for Mental Health for Children and Adolescents in Denmark.
- Two additional models are scheduled for announcement in October.
Connected Companies & Entities
“They later joined Antler's matchmaking programme, where they met Sally.”
Ontology Mapping & Concepts
Related Market Signals & Shifts
Recent verified developments and strategic activity across this market segment.
Synthesia launches Roleplay Sessions for live AI coaching
British startup Synthesia launched Roleplay Sessions, an interactive enterprise training product that lets employees practice conversations (sales pitches, performance reviews, customer complaints) with an AI avatar that responds, pushes back, and scores performance against a rubric. Roleplay is the first release under a broader Sessions platform that Synthesia plans to expand into formats such as job interviews and candidate screening. The company combines proprietary avatar and voice technology with underlying reasoning from OpenAI, and adds rubrics, performance data and analytics to position itself as a performance-management platform. Early customers include large, scaled commercial rollouts with major European and Fortune 100 enterprises. Synthesia says the product will broaden to smaller customers as inference costs fall.
Study: AI Builds Trust Better Than Humans, Enabling Scams
A recent study found that a language model (an older Claude model from Anthropic) was more effective than human interlocutors at building exploitable trust over week-long conversations. In the experiment, 46% of participants installed an app recommended by the AI versus 18% after conversations with humans; many participants did not realize they were talking to an AI. Researchers warn generative AI could automate the slow, resource-intensive relationship-building phase of 'pig butchering' scams, increasing the threat of consumer and corporate social-engineering attacks. Anthropic says newer model versions include added safety mechanisms. The article also references recent incidents involving OpenAI and an attack on Hugging Face as evidence of risks from powerful autonomous AI behavior.
Thinking Machines builds AI that listens while it talks
Thinking Machines Labs, founded by former OpenAI CTO Mira Murati, has published a technical preview describing an "interaction model" designed for full‑duplex, real‑time multimodal conversation. The model targets subsecond latency (about 0.4 seconds) so it can respond while a person is still speaking — including interrupting to correct errors, provide proactive answers, or perform simultaneous translation — and supports continuing conversation while working on background tasks. Company demos show these capabilities but the model is not yet public; Thinking Machines plans a limited research preview for testers in the coming months. Related technical disclosures (research preview) describe TML‑Interaction‑Small (a mixture‑of‑experts architecture) and benchmark wins on audio/vision timing evaluations, indicating the work is positioned as a research/technical preview rather than a commercial product.
Receive all future signals for Antler & competitors in your briefing
Never miss material M&A deals, leadership changes, account wins, or product rollouts. Polaris7 monitors 20,000+ primary sources 24/7, delivering verified telemetry directly into your executive briefing.
