Tech startup Human AI Labs has launched Blind Spot, a Romanian tool designed to assess whether websites are ready for the next stage of online search, in which artificial intelligence systems increasingly extract, summarize and recommend information directly to users rather than simply displaying links.

According to the company, the platform evaluates a website's digital infrastructure against 70 deterministic criteria, covering AI crawler accessibility, structured data, semantic clarity, citable content, digital authority and trust signals.

Blind Spot then generates an AI visibility score, identifies issues that may make a website difficult for generative AI systems to understand or cite, and provides concrete recommendations for addressing them.

'I have worked in IT for more than 25 years and have experienced all the major transitions of the internet, from static websites to e-commerce, cloud, mobile and automation. But the change brought by AI to search is different, because of the way information is understood, selected and presented. Blind Spot came from a simple question: does AI know who you are, what you do and why it should recommend you?,' said Human AI Labs founder Andrei Raileanu.

The launch comes as search behavior is changing rapidly, with users increasingly turning to platforms such as ChatGPT, Perplexity, Gemini, Claude and Google AI Overviews for direct answers rather than browsing traditional search results.

For companies, this means visibility increasingly depends not only on Google rankings but also on whether their websites can be correctly understood, extracted and cited by AI systems.

'For years, companies have optimized for Google: keywords, backlinks, titles, meta tags and so on. All of these remain important, but AI search works differently. It needs structure, clarity, coherent data, easily citable content and authority signals that it can verify. A website can be flawless for humans and still be difficult for AI to read,' Raileanu added.

During its beta phase, Blind Spot audited several Romanian websites, producing an average AI visibility score of 48 out of 100. The most common issues included incomplete structured data, difficult-to-cite content, unclear authority signals and an inconsistent presence across sources that AI systems may interpret as references.

Unlike tools that monitor how brands appear in AI-generated answers, Blind Spot examines the underlying website infrastructure and identifies what needs to be fixed to make a site easier for AI systems to understand, cite and recommend. Its score is calculated deterministically, without AI being involved in the assessment, while reports include identified problems and recommendations. For certain technical criteria, the platform can also provide code examples and implementation guidance.

The Blind Spot architecture, including its normalized extraction methodology, deterministic scoring and semantic gap analysis, is the subject of a patent application filed with Romania's State Office for Inventions and Trademarks (OSIM) in June 2026.

'We don't want to sell panic around AI; we want to provide clarity. When AI misinterprets a website, the first step is to see where the chain breaks: what is missing, what is unclear and what can be improved. Blind Spot provides that diagnosis,' Raileanu said.

Blind Spot is available through myblindspot.eu, with several access options. The platform also offers a programme giving eligible NGOs six months of free access and a startup programme including a free 'Diagnostic' audit.

Human AI Labs plans to add tools for analyzing brand perception in AI models, monitoring discrepancies between a brand's intended messaging and how AI describes it, and providing dedicated tools for partners and agencies.

These developments complement the technical audit currently available and will allow users to track over time how implemented changes affect their visibility in AI-generated results.

Human AI Labs develops products and methodologies that help organizations adapt to changes brought by artificial intelligence in search, content and digital infrastructure.

Founded by Andrei Raileanu, an IT specialist with more than 25 years of experience in developing business-critical applications and digital products, the company combines technical expertise, applied research and an understanding of how AI is changing access to information. AGERPRES (RO - writing by: Oana Tilica; EN - writing by: Simona Klodnischi)