AI Companies Promoting Transparency Cannot Even Be Honest With Each Other Report Finds

Despite Marketing Themselves as Open and Transparent Major AI Firms Fail to Meet Their Own Standards\n\nCompanies that market themselves as champions of AI transparency are failing to be honest even with each other, according to a new investigation by The Verge.\n\n### The Problem\n\n- AI companies positioning themselves as pro-transparency are not being honest internally\n- Industry claims of openness are contradicted by actual practices\n- Self-regulation is proving inadequate for the AI sector\n\n### What This Means\n\nThe AI industry has made transparency a key marketing differentiator. Companies like Anthropic, OpenAI, and others frequently discuss the importance of open research and honest communication about AI capabilities and risks. But behind the scenes, the competitive pressure to maintain advantages leads to selective disclosure and strategic opacity.\n\n### The Trust Gap\n\nThis creates a dangerous trust gap: policymakers, researchers, and the public are asked to trust AI companies' self-reported safety claims, but if companies cannot even be honest with each other, those claims become unreliable.\n\n### Implications\n\nAs governments worldwide draft AI regulations, the transparency gap raises fundamental questions about whether self-regulation can work and whether external auditing mechanisms are needed.\n\nSource: The Verge, Jess Weatherbed