About AI Aware
Building leading artificial intelligence content detectors for commercial, public and academic sectors.
- Founded in London in 2023
- Funded by Innovate UK
- Research with City St George’s, University of London
- Member of the Content Authenticity Initiative
Company story
AI Aware began in 2023, when Stephen Harmston and Dr Tillman Weyde combined commercial data experience with academic research into multi-modal AI. The company now detects AI-generated text, images, video and audio from one platform.
AI Aware (the trading name of Happy AI Ltd) is a deep tech AI content detection company, headquartered in London, UK.
Founders
Commercial and technical expertise, with over 20 years of AI experience.
Stephen Harmston
Stephen Harmston is leading on product development and commercialisation. He created and successfully exited VentureOne Europe, a venture-backed database business. Formerly Founder and MD at Discover, an AI-driven entertainment platform and Head of data reporting at YouGov Plc, where he built a data reporting business to £multi-million turnover.
Dr Tillman Weyde
Tillman is leading on AI Aware’s technical approaches to AI content detection, building on his published multi-modal AI research. Dr Weyde is Reader in Computer Science and Director of the Artificial Intelligence Research Centre at City St George’s, University of London. He has worked in machine learning for more than 20 years and has published over 100 research papers.
R&D funding from Innovate UK
Technology grants from Innovate UK, the United Kingdom’s innovation agency, have funded our AI development work over the past 2 years. Only around 5 per cent of companies applying for such funding are supported – the main requirements for funding are innovation level (as judged by a panel of experts) and prospects for commercialisation.
Detecting AI-generated text, images, audio and video
Our innovative approach to AI text, video, audio and image detection enables very high rates of accuracy compared to other approaches as the more information used to detect AI, the better the detection. Our approach is based on maximising information.
We diverge from current AI detection methods that just use Machine Learning. These detect patterns consistent with AI and have proven to be inaccurate, particularly with types of content that they have not seen before.
We combine AI detection models using a unique cross-modal approach, detecting content with reverse AI, statistical patterns, and a range of signals that people use. AI Aware combines these techniques with additional metadata to form a unique proprietary system.
This approach stems from research by one of our co-founders, Dr Tillman Weyde, who is Reader in Computer Science at City St George’s, University of London.
We have tested the AI Aware approach for text detection against other companies and has demonstrated that it is more accurate. See the AI text detection benchmarks.
The AI Aware content detection platform is neutral. We don’t judge whether AI-generated content is ‘good’ or ‘bad’. Instead, we give users easy-to-understand visual guides. These show where AI has been used, where adapted and where human based. Importantly, we provide calibrated probabilities indicating the confidence of AI being detected rather than a binary yes/no.
Over 20 years of artificial intelligence experience
The founders have a mix of commercial and technical expertise having built data companies worth over £20m and a range of AI-based products including video content recommendation systems, AI video scheduling and AI content detection. You can learn more about the methodology of AI Aware and why it matters here and about the team.
See what AI Aware finds
Run 3 free checks now, no sign-up needed.
Try it free, no sign-up