Grammarly recently disabled its AI 'Expert Review' feature following a class-action lawsuit led by journalist Julia Angwin. The feature, which offered writing suggestions styled after real authors and academics, faced criticism for using individuals' names and likenesses without their explicit consent or compensation. Angwin's lawsuit argues this practice violates New York and California laws regarding commercial use of identity, prompting Grammarly's parent company, Superhuman, to reimagine the feature with greater expert control.
The broader implications of AI's unpredictable nature are also surfacing in security and safety discussions. Cloudflare now offers its AI Security for Apps service, providing a crucial layer for AI-powered applications and free AI discovery to combat 'shadow AI' deployments. Meanwhile, Wiz has developed new capabilities to analyze AI systems across infrastructure, model, data, and application layers, addressing complex interconnected threats. On the safety front, Amazon's Alexa Plus demonstrated a critical flaw by suggesting a dangerous mix of bleach and vinegar for mold removal, underscoring the vital need to verify AI-generated advice, especially for health and safety.
AI's influence extends into education, where college professors observe a decline in student reading engagement, partly due to AI's summarization capabilities and a preference for bite-sized media. Socially, a TikTok trend highlights individuals using AI to justify bad behavior, while teenagers are creating 'slander pages' targeting teachers with AI tools like Viggle AI, prompting platforms like Meta and TikTok to remove violating content. Amid these challenges, innovation continues with Looking Glass launching Musubi, an AI-powered holographic frame. CEO Shawn Frayne emphasizes Musubi's local AI processing to protect user privacy, offering a 7-inch frame that displays 3D photos and videos without cloud reliance.
Ethical considerations are also at the forefront, as seen with Microsoft backing AI firm Anthropic in its dispute with the Pentagon. Anthropic seeks assurances against its AI being used for autonomous weapons or mass surveillance, highlighting a critical debate over AI's military applications. Historically, the automation of tasks, like player pianos replacing pianists, shows that while technology evolves, the demand for human artistry and connection persists. Books like 'AI Shepherds and Electric Sheep' further explore AI's impact on faith and humanity, encouraging discernment and a balanced perspective on technology's role in our lives.
Key Takeaways
- Grammarly disabled its AI 'Expert Review' feature and faces a class-action lawsuit from journalist Julia Angwin for using individuals' likenesses without consent.
- Cloudflare launched its free AI Security for Apps service and AI discovery to secure AI-powered applications and detect 'shadow AI'.
- Wiz developed new capabilities to analyze AI systems across infrastructure, model, data, and application layers to identify interconnected threats.
- Amazon's Alexa Plus provided dangerous advice, suggesting mixing bleach and vinegar, highlighting the need to verify AI-generated safety information.
- College professors note a decline in student reading engagement, partly attributed to AI tools and changing media consumption habits.
- Teenagers are using AI tools like Viggle AI to create 'slander pages' targeting teachers on platforms like TikTok and Instagram, leading to content removal by Meta and TikTok.
- A TikTok trend is emerging to call out individuals who use AI to justify negative or irresponsible behavior in relationships.
- Looking Glass launched Musubi, an AI-powered holographic frame, with CEO Shawn Frayne emphasizing local AI processing for user privacy.
- Microsoft supports AI firm Anthropic in its dispute with the Pentagon, seeking to block a ban and negotiate terms regarding AI use for autonomous weapons or mass surveillance.
- Historical parallels with automation, like player pianos, suggest that while AI automates tasks, demand for human artistry and connection persists.
Grammarly stops AI expert cloning after backlash and lawsuit
Grammarly has decided to disable its AI 'Expert Review' feature after facing criticism and a class-action lawsuit. The feature offered writing suggestions inspired by real authors and academics without their explicit consent. Grammarly's parent company, Superhuman, stated they will reimagine the feature to give experts control over their representation. Journalist Julia Angwin is leading the lawsuit, arguing that using individuals' names and likenesses for profit without permission is illegal. The company apologized for missing the mark and plans to develop the feature differently.
Grammarly sued over AI feature impersonating authors
Grammarly is facing a class-action lawsuit for its AI 'Expert Review' feature, which presented editing suggestions as if they came from real authors and academics without their permission. The feature, now shut down, led to journalist Julia Angwin filing a suit arguing that Grammarly profited from misusing individuals' identities. Superhuman, Grammarly's parent company, stated they are disabling the feature to allow experts more control. The lawsuit claims this practice violates laws in New York and California regarding the commercial use of a person's name and likeness. Angwin expressed surprise at being included and criticized the quality of the AI-generated advice.
Lawsuit filed against Grammarly for AI impersonating writers
Journalist Julia Angwin has filed a class-action lawsuit against Grammarly for using her identity in its 'Expert Review' AI editing feature without her consent. This feature, still in beta, provided writing suggestions based on the styles of well-known authors and journalists. Angwin discovered her likeness was used when The Verge tested the feature, which also included other Verge staff members. Grammarly did not inform or ask permission from the individuals whose identities were used.
Grammarly halts AI feature after lawsuit over cloning voices
Grammarly has stopped its AI 'Expert Review' feature, which mimicked real authors and academics without their consent, on the same day it was sued. This feature offered writing suggestions styled after recognized experts, but these individuals had not agreed to have their names or expertise used by the AI. The lawsuit argues that Grammarly commercially exploited these professionals' reputations without permission or compensation. This case highlights the legal challenges surrounding AI companies using personal likeness and professional identity.
New security approach needed for interconnected AI systems
Modern AI applications create complex risks by connecting multiple components across different layers: infrastructure, model, data, and application. Traditional security tools struggle to identify these interconnected threats. For example, a chatbot might seem harmless, but if its infrastructure is exposed, its model has weak guardrails, its data access is broad, and its application tools are powerful, it creates a significant security risk. Understanding how these layers interact is crucial for identifying and mitigating AI risks effectively. Wiz has developed new capabilities to analyze AI systems across these layers, providing security teams with a unified view of AI security posture.
Cloudflare launches free AI security for all apps
Cloudflare has made its AI Security for Apps service generally available, offering a security layer for AI-powered applications regardless of their model or hosting provider. The company is also offering AI discovery for free on all plans to help teams find and secure 'shadow AI' deployments. New features include custom topic detection and expanded collaboration with IBM and Wiz. AI-powered applications present unique security challenges due to their unpredictable nature, making traditional security rules insufficient. Cloudflare's tool aims to automatically identify AI endpoints and detect threats like prompt injection.
Colleges see students reading less amid AI and changing media habits
College professors are noticing a decline in students' engagement with assigned reading, a trend attributed partly to the rise of AI and changing media consumption habits. Students are reportedly less inclined to do the reading before class and participate in discussions about it. Experts suggest that the preference for bite-sized information from video and audio content, along with AI's ability to summarize texts or write essays, contributes to this shift. This presents a challenge for higher education, impacting critical thinking and overall learning. Educators are exploring new methods, like interactive discussions and multimedia resources, to make reading more engaging and relevant.
New frame brings AI-powered holographic photos to your home
Looking Glass has launched Musubi, an AI-powered holographic frame designed to display photos and videos in 3D. Unlike headset-based AR, Musubi is a 7-inch frame that processes images locally on a PC or MacBook, with no Wi-Fi or subscription needed. It can store up to 1,000 images or 30-second videos. CEO Shawn Frayne emphasizes the use of local AI to protect user privacy, ensuring images are not sent to large tech companies. This device builds on Looking Glass's Hololuminescent display technology, offering a more accessible entry into holographic experiences.
Automation's impact on human artistry explored
The automation of the pianist job by player pianos over 130 years ago serves as a historical parallel to current concerns about AI replacing human artists. Despite technological advancements, human musicians continue to thrive, with job opportunities growing over time. This is because consumer demand values the human element in art, not just the final product. The article suggests that while AI may automate certain tasks, the demand for human connection and unique performances ensures the continued relevance of artists across various fields, from music to other services.
TikTok trend calls out AI use for bad behavior
A new trend on TikTok is highlighting how people use artificial intelligence to justify their negative actions, particularly in relationships. Users create videos exposing instances where AI is employed as an excuse for hurtful or irresponsible behavior. This trend serves as a counter-narrative to the idea that AI can absolve individuals of personal responsibility. It reminds viewers that while AI is a powerful tool, it should not be used to excuse harmful conduct, emphasizing the importance of human connection and accountability in the digital age.
Book reviews AI's impact on faith and humanity
The book 'AI Shepherds and Electric Sheep: Leading and Teaching in the Age of Artificial Intelligence' by Sean O'Callaghan and Paul A. Hoffman offers a balanced perspective on AI for Christian leaders and educators. It avoids both uncritical acceptance and reflexive rejection of AI, instead focusing on understanding its concepts and implications. The book explores what it means to be human in light of technology, emphasizing the 'imago Dei' and embodiment. It provides frameworks for faithful responses to AI, addressing pastoral questions about its use in ministry and technology's impact on relationships, encouraging discernment.
Teens use AI tools to create 'slander pages' targeting teachers
Teenagers are using AI tools like Viggle AI to create 'slander pages' on social media platforms like TikTok and Instagram, targeting their teachers with mocking content. These pages employ slang and sometimes extremist symbols, with videos featuring teachers' faces superimposed onto other content, including disturbing scenarios. Platforms like Meta and TikTok have removed some violating content and are implementing rules to catch similar posts. School districts are aware of the issue and state that students responsible will face disciplinary and potential legal consequences.
AI can give dangerous advice, experts warn
Artificial intelligence assistants, while often helpful, can sometimes provide incorrect or dangerous advice, as discovered when asking about removing mold from washing machines. Amazon's Alexa Plus suggested mixing bleach and vinegar, which creates toxic chlorine gas, without issuing a warning. While other AI tools provided safer advice, Alexa's response remained risky even after being informed of the danger. This highlights that AI does not truly 'know' things but generates responses based on data patterns, making it crucial to double-check AI-generated information, especially concerning health, safety, or repairs.
Microsoft backs AI firm Anthropic against Pentagon ban
Microsoft is supporting AI company Anthropic in its dispute with the Pentagon, urging a court to temporarily block the military's ban on Anthropic's products. Microsoft argues this would allow for an 'orderly transition' and prevent disruption to the military's use of advanced AI. The ban followed failed contract negotiations, with Anthropic seeking assurances against using its AI for autonomous weapons or mass surveillance, while the DOD wanted unfettered access. Microsoft believes a temporary block will help Anthropic and the DOD find a negotiated resolution that serves all parties and avoids significant business impacts.
Sources
- Grammarly says it will stop using AI to clone experts without permission
- Grammarly Is Facing a Class Action Lawsuit Over Its AI ‘Expert Review’ Feature
- One of Grammarly’s ‘experts’ is suing the company over its identity-stealing AI feature
- Grammarly Faces Class Action Over AI 'Expert' Impersonation
- Reducing AI Risk Across Modern AI Applications
- AI Security for Apps is now generally available
- Did anybody do the reading? Colleges grapple with a generational shift in learning
- This Digital Picture Frame Wants to Bring People Closer to a Holographic Future
- Play It Again, Claude
- People Use AI to Justify Bad Behavior. This TikTok Trend Calls Them Out
- Shepherding Souls in the Age of Artificial Intelligence: A review of AI Shepherds and Electric Sheep
- Teens Are Using AI-Fueled ‘Slander Pages’ to Mock Their Teachers
- What The Tech: Can artificial intelligence be wrong?
- Microsoft says court should temporarily block Pentagon's blacklist of Anthropic
Comments
Please log in to post a comment.