\n
![](\"https:\/\/content.fortune.com\/wp-content\/uploads\/2024\/03\/AP24066708384190-e1709755520852.jpg?w=2048\")
<\/p>\n [ad_1]
\n<\/p>\n
A Microsoft<\/a> engineer is sounding alarms about offensive and harmful imagery he says is too easily made by the company\u2019s\u00a0artificial intelligence image-generator tool<\/a>, sending letters on Wednesday to U.S. regulators and the tech giant\u2019s board of directors urging them to take action.<\/p>\n Shane Jones told The Associated Press that he considers himself a whistleblower and that he also met last month with U.S. Senate staffers to share his concerns.<\/p>\n The Federal Trade Commission confirmed it received his letter Wednesday but declined further comment.<\/p>\n Microsoft said it is committed to addressing employee concerns about company policies and that it appreciates Jones\u2019 \u201ceffort in studying and testing our latest technology to further enhance its safety.\u201d It said it had recommended he use the company\u2019s own \u201crobust internal reporting channels\u201d to investigate and address the problems. CNBC was first to report about the letters.<\/p>\n Jones, a principal software engineering lead, said he has spent three months trying to address his safety concerns about Microsoft\u2019s Copilot Designer, a tool that can generate novel images from written prompts. The tool is derived from another AI image-generator, DALL-E 3, made by Microsoft\u2019s close business partner OpenAI.<\/p>\n \u201cOne of the most concerning risks with Copilot Designer is when the product generates images that add harmful content despite a benign request from the user,\u201d he said in his letter addressed to FTC Chair Lina Khan. \u201cFor example, when using just the prompt, \u2018car accident\u2019, Copilot Designer has a tendency to randomly include an inappropriate, sexually objectified image of a woman in some of the pictures it creates.\u201d<\/p>\n Other harmful content involves violence as well as \u201cpolitical bias, underaged drinking and drug use, misuse of corporate trademarks and copyrights, conspiracy theories, and religion to name a few,\u201d he told the FTC. His letter to Microsoft urges the company to take it off the market until it is safer.<\/p>\n This is not the first time Jones has publicly aired his concerns. He said Microsoft at first advised him to take his findings directly to OpenAI, so he did.<\/p>\n He also publicly posted a letter to OpenAI on Microsoft-owned LinkedIn<\/a> in December, leading a manager to inform him that Microsoft\u2019s legal team \u201cdemanded that I delete the post, which I reluctantly did,\u201d according to his letter to the board.<\/p>\n In addition to the U.S. Senate\u2019s Commerce Committee, Jones has brought his concerns to the state attorney general in Washington, where Microsoft is headquartered.<\/p>\n Jones told the AP that while the \u201ccore issue\u201d is with OpenAI\u2019s DALL-E model, those who use OpenAI\u2019s ChatGPT to generate AI images won\u2019t get the same harmful outputs because the two companies overlay their products with different safeguards.<\/p>\n \u201cMany of the issues with Copilot Designer are already addressed with ChatGPT\u2019s own safeguards,\u201d he said via text.<\/p>\n A number of impressive AI image-generators first came on the scene in 2022, including the second generation of OpenAI\u2019s DALL-E 2. That \u2014 and the subsequent release of OpenAI\u2019s chatbot ChatGPT \u2014 sparked public fascination that put commercial pressure on tech giants such as Microsoft and Google<\/a> to release their own versions.<\/p>\n But without effective safeguards, the technology poses dangers, including the ease with which users can generate\u00a0harmful \u201cdeepfake\u201d images<\/a>\u00a0of political figures, war zones or nonconsensual nudity that falsely appear to show real people with recognizable faces. Google has temporarily suspended its\u00a0Gemini chatbot\u2019s<\/a>\u00a0ability to generate images of people following outrage over how it was depicting race and ethnicity, such as by putting people of color in Nazi-era military uniforms.<\/p>\n<\/div>\n\nSubscribe to the Eye on AI newsletter to stay abreast of how AI is shaping the future of business. Sign up<\/a> for free.<\/div>\n