Trust in AI has become the industry’s biggest liability. AI companies say they take user privacy seriously, but somehow none will tell you exactly what that means.
To help find an answer, Cybernews has launched the AI Trustworthiness Ranking, a new project assessing 500 AI companies from 36 countries on how responsibly they handle security, data privacy, organisational transparency, and public perception. The project’s mission is to cut through the noise of a crowded AI market, where it is often unclear how the companies behind popular tools actually operate, or what comes of user data once someone starts using the product.
Each company receives a Trustworthiness Score between 0 and 100, calculated from the four weighted categories. Those scoring 75 or above earn the title of ‘AI Trustworthiness Leader of 2026’. The ranking spans 21 different categories, from assistants and productivity software to coding, music, and creative tools.
The stakes are rising as AI systems are moving beyond simple question-answering capabilities. Dr. Akshika Wijesundara, an advisory board member, and Co-Founder of Snapdrum and TAI Labs, argues that the shift towards autonomous AI Agents raises the bar for corporate accountability.
“AI is shifting from answering questions to taking actions, reading inboxes, moving money, and making decisions on our behalf,” Dr. Wijesundara commented. “A chatbot that mishandles data is a privacy problem. An Agent with broad permissions and weak governance is a security problem, with mistakes propagating through real systems at machine speed. Companies that clearly show how they protect data, use it, and govern what their AI is allowed to do will have a meaningful advantage in earning long-term trust.”
Who ranks on top?
Google’s Gemini took the number one spot overall. According to the ranking, the ten most trustworthy AI companies are Google (Gemini), Krisp, Fireflies.ai, Adobe, Magnific, Writesonic, Veryfi, Salesforce, Grammarly, and Lovable.
By category, Office & Productivity companies scored highest on average, at 78 out of 100, while Music & Audio companies scored the lowest, averaging at just 54. The top five most trustworthy categories are Office & Productivity (78), Business Management (75), Coding & Development (71), AI assistants (69), and Writing & Editing (68). The lowest-scoring categories are Interior & Architectural Design (59), Business Research (58), Art & Creative Design (57), AI Detection & Anti-detection (57), and Music & Audio (54).
Of the four pillars assessed, security received the lowest average score at just 32 out of 100, while organisational transparency was the strongest-performing pillar, with an average score of 90 out of 100. On average public perception and privacy scored 71 and 70, respectively.
The ranking also found that trustworthiness, according to its criteria, grows with company size. On average, companies with 1-10 employees had a score of 56, while larger companies with 5000+ employees scores 82.
According to Voldemaras Kadys, member of the AI Trustworthiness Ranking Advisory Board and Head of Security and Platform Engineering at Mediatech, security and privacy aren’t just about legal compliance – they’re also about reputation and trust: “AI tools can have access to huge amounts of sensitive user and business data, so companies need to be clear about how they use, store, and share that data. They must also show that they have strong security measures in place to protect it. If a company isn’t transparent about its data practices, or suffers a security incident, it can quickly lose the trust of its customers – and that trust isn’t easy to win back,” says Kadys.
The big takeaway: most AI companies lack clear disclosures on AI training
To assess the Data Privacy pillar, the company reviewed the privacy policies of all 500 companies, focusing on how clearly each one explains its data collection, sharing, use, and retention practices.
The findings point to a significant transparency gap. Nearly two-thirds (63%) of companies don’t clearly disclose whether they use customer data to train their AI models. Within that group, 42% say nothing about AI training on user data in their privacy policies at all, while 21% address it only in vague terms.
Data retention practises show a similar pattern. Sixty-five percent of companies fail to state how long they keep user data: 9% make no mention of retention or deletion at all, and 56% touch on it only vaguely, without specifying anything solidly.
The findings suggest that while AI companies have gotten better at presenting themselves more transparently, the technical safeguards and disclosures that directly impact users, still lag behind.
