Events
Categories
Programs
Events
Categories
Programs
Microsoft informant alerts US authorities that OpenAI-supported Copilot can effortlessly generate destructive visuals
Similar to Google’s Gemini AI image generator, Microsoft’s Copilot, backed by OpenAI’s ChatGPT, can also produce highly concerning yet realistic counterfeit visuals. An informant has alerted US officials and Microsoft’s board by sending letters warning them about this problem.
A Microsoft employee has expressed worries about damaging and inappropriate images produced by the company's AI image-creation software. Shane Jones, who views himself as a truth-teller, has written to American regulatory bodies and the board of directors at Microsoft, encouraging them to intervene, as indicated in a news report by the Associated Press.
Jones had a recent meeting with staff members of the US Senate to talk about his worries and he also wrote a letter to the Federal Trade Commission (FTC). The FTC acknowledged receiving the letter but chose not to give any more comments.
Microsoft expressed its dedication to tackling employee issues and valued Jones' contribution to the technology testing. Nevertheless, it suggested utilizing internal communication routes to probe and rectify the problems.
Jones, a chief software engineering manager, has devoted a quarter of a year to tackling security issues associated with Microsoft's Copilot Designer. He underscored the potential danger of the tool producing damaging content, even in response to harmless instructions. To illustrate, if the tool is given a cue like 'car accident,' it might generate unsuitable, sexually suggestive imagery involving women.
Jones underscored to FTC Chair Lina Khan that Copilot Designer can potentially create damaging content, even in response to benign user inputs. For instance, it may sometimes produce inappropriate sexualized images of women in response to a 'car accident' prompt. He also brought to attention other troubling content such as violent scenes, partisan bias, minor alcohol consumption and drug abuse, violations of copyright, unfounded conspiracy theories, and religious symbols.
Jones has earlier voiced these worries in public. Suggested by Microsoft to contact OpenAI, he presented his discoveries to them. In December, he published a letter on LinkedIn directed towards OpenAI, which resulted in Microsoft's legal department insisting on its removal. Regardless, Jones has continued, taking his issues to the US Senate's Commerce Committee and the Office of the Attorney General in Washington State.
Jones emphasized that the primary concern is with OpenAI's DALL-E model. However, users of OpenAI's ChatGPT for AI image generation are less prone to come across damaging results because of the various safety precautions that the two firms have put in place.
He communicated through text that a lot of the issues associated with Copilot Designer are already addressed by the inherent protective measures of ChatGPT.
The arrival of remarkable AI image creators in 2022, such as OpenAI's DALL-E 2 and the later launch of ChatGPT, sparked considerable attention from the public. This led major technology companies like Microsoft and Google to create their own editions.
Nonetheless, absent strong protective measures, the technology poses dangers, permitting users to generate damaging "deepfake" pictures portraying political leaders, war scenarios, or involuntary nudity, incorrectly associating them with actual people.
Following worries, Google momentarily halted the image creation function of the Gemini chatbot. This was primarily due to disputes over representations of race and ethnicity, including situations where individuals of color were depicted in military uniforms from the Nazi era.
(Incorporating information from various sources)
Search for us on YouTube
Top-Rated Programs
Associated Reports
NVIDIA's Jensen Huang believes AI hallucinations can be resolved and anticipates artificial general intelligence to be achieved in about 5 years.
OpenAI's Sora has the capability to create lifelike nude clips, prompting developers to quickly find a solution.
Apple has at last unveiled MM1, its combined AI framework for generating text and images.
Microsoft has recruited Mustafa Suleyman, cofounder of DeepMind, to head their new customer AI group.
NVIDIA's Jensen Huang is confident that AI hallucinations can be tackled and expects general artificial intelligence to be available in roughly 5 years.
OpenAI's Sora has the ability to produce convincing nude films, causing developers to hasten in finding a remedy.
Apple has ultimately released MM1, its integrated AI system for text and image production.
Microsoft has engaged Mustafa Suleyman, the cofounder of DeepMind, to guide their fresh consumer AI division.
Find us on YouTube.
Firstpost holds the exclusive rights, protected by copyright law, until the year
+ There are no comments
Add yours