Seattle, Washington – A Microsoft AI engineer has raised concerns about the company’s AI text-to-image generator, Copilot Designer, warning that it may be generating violent and sexual imagery. Despite repeated warnings from the engineer, Shane Jones, Microsoft has allegedly failed to take action to address these issues.
Jones, who participated in red-teaming efforts to test the tool’s vulnerabilities, claims that Microsoft ignored his warnings and did not implement safeguards or disclose the mature content rating of the product. Instead of addressing the concerns, Microsoft reportedly referred Jones to report the issue to OpenAI, the developer of the DALL-E model that powers Copilot Designer.
When OpenAI failed to respond to the reported issues, Jones took it upon himself to alert the public by posting an open letter on LinkedIn and sending letters to lawmakers, stakeholders, and regulatory bodies such as the Federal Trade Commission. In his communication to the FTC, Jones expressed his concerns that Microsoft and OpenAI were aware of the issues but continued to market the product without intervention.
According to reports, Jones also highlighted the potential risks associated with Copilot Designer, including the generation of inappropriate and harmful content such as sexually objectified images, political bias, underage drinking, conspiracy theories, and more. He urged Microsoft’s board of directors to conduct an independent review of the company’s AI decision-making processes and responsible incident reporting.
While Microsoft has not confirmed whether steps are being taken to filter images, attempts to replicate prompts shared by Jones resulted in error messages. The company provided a statement emphasizing their commitment to addressing employee concerns and enhancing the safety of their technology. OpenAI, on the other hand, did not respond to requests for comment on the matter.
The controversy surrounding Microsoft’s Copilot Designer highlights the ongoing challenges and ethical considerations in the development and deployment of AI technologies. It underscores the importance of robust safety measures and oversight to mitigate potential risks and ensure responsible AI practices in the industry.