AI Scraping Opt-Out Bypassed by OpenAI and Anthropic, ABC Tells Parliament

The article argues that relying on the robots.txt protocol as a mechanism for copyright compliance is fundamentally flawed because it is a voluntary convention with no enforcement power. Major AI companies like OpenAI and Anthropic have been documented bypassing these "opt-out" signals, demonstrating that the onus cannot fairly remain on publishers to protect their content. This technical reality undermines proposals for AI industry self-regulation and highlights the inadequacy of opt-in or opt-out frameworks in ensuring genuine consent or remediation for unauthorized data usage. Consequently, the author contends that existing copyright laws, which mandate permission before use, offer the only viable legal remedy. There is a pressing need to close regulatory gaps by extending Australia’s News Bargaining Incentive to include AI companies, similar to how social media platforms are currently regulated. Without such legislative updates, AI firms continue to exploit content without compensation, leveraging their technical capacity to ignore standard web protocols while facing minimal legal or financial consequences for infringement. This issue is critical to open data and copyright discourse because it challenges the assumption that technical safeguards can substitute for legal rights. It illustrates the dangerous asymmetry where creators must constantly monitor and defend their data, while large technology entities face negligible barriers to bypassing protections. The outcome in Australia serves as a global test case, signaling whether strong copyright enforcement can successfully compel AI developers to license training data, thereby setting a precedent for balancing technological innovation with the rights of content creators.

Source: techtimes.com
Published on 2026-10-07

AI Scraping Opt-Out Bypassed by OpenAI and Anthropic, ABC Tells Parliament

Related news