🗞 OpenAI Rolls Out ChatGPT Edu and Discounted Access for Nonprofits Amid Controversies 🎓

🔎 The Latest on the AI Frontier:

• OpenAI Offers Discounted ChatGPT Access for Education and Nonprofits 🎓

• OpenAI Combats Influence Operations with AI Tools 🕵️

• Instagram's Deceptive Design Hinders AI Data Training Opt-Out 😾

• Apple to Unveil AI-Powered Features in iOS 18 at WWDC 🍎

• McKinsey and Microsoft Join Forces to Drive Business Value with Gen AI 🤝

• Survey: Americans Anticipate AI Abuses in 2024 Election 🗳️

• Study Exposes Ease of Voice Cloning Political Figures on AI Platforms 🎙️

🎓 OpenAI introduces ChatGPT Edu for universities and discounted access to ChatGPT for nonprofits. Read More

• ChatGPT Edu, built on the faster GPT-4o model, allows universities to use AI for reviewing resumes, writing grant applications, and assisting with grading at an affordable rate.

• OpenAI for Nonprofits offers ChatGPT Team at $20 per month per user, while larger organizations can get ChatGPT Enterprise at a 50% discount.

• The announcements come amid recent resignations of key researchers citing safety concerns and criticism of CEO Sam Altman for rolling out an AI voice resembling Scarlett Johansson without her consent.

🔎 OpenAI detects and removes influence operations from Russia, China, Iran, and Israel using its AI tools. Read More

• OpenAI's report reveals that bad actors used tools like ChatGPT to generate social media comments, fake account names and bios, and debug code for their influence operations.

• Despite using AI to produce more content and make fewer errors, these operations struggled to gain significant traction or reach large audiences, with some only receiving engagement from other fake accounts.

• OpenAI banned accounts linked to 5 covert influence operations in the past 3 months, including Russia's Doppelganger and China's Spamouflage, which are known for targeting Ukraine and pushing pro-China messages, respectively.

😠 Instagram is using deceptive UX to make it nearly impossible for users to opt out of having their data used to train AI models. Read More

• Meta has started mining Instagram images and videos to train its AI models, effectively killing the platform's spirit and exploiting creators against their will.

• While EU residents can technically opt out due to privacy laws, Meta employs deceptive design tactics to dissuade and discourage users from doing so, such as confusing legal jargon, extra steps, and implying that objections may not be honored.

• Instead of a simple "YES or NO" option, the opt-out process involves navigating through misleading text, providing explanations, and completing authentication steps, making it unnecessarily difficult for users to exercise their rights under EU law.

🤖 Apple is expected to announce a slew of new AI features as part of iOS 18 at the upcoming WWDC keynote. Read More

• Siri is set to get generative AI smarts, thanks to a deal with OpenAI to incorporate ChatGPT tech, enabling smarter understanding of commands, spoken answers, awareness of user context, and the ability to summarize information and interact with apps.

• AI-powered photo editing capabilities are anticipated, including object removal, text-based editing commands like "make the sky more blue," and the ability to identify and edit specific photo elements like eyes, hair, and clothes.

• Other possible AI features include the creation of custom emoji based on what users are texting, smarter Spotlight searches that can dig deep into apps, AI-assisted coding in Xcode, AI-powered health & fitness coaching, and AI-enhanced journaling to detect user moods.

🤝 McKinsey and Microsoft collaborate to drive business value creation with generative AI (gen AI) agents enabled by Copilot Studio. Read More

• The alliance will offer an end-to-end transformation approach, helping businesses re-imagine core workflows with Microsoft's Copilot Studio and ensuring adoption of tools and processes to improve service quality, enable growth, and drive efficiency.

• Gen AI enterprise use cases could yield $2.6 trillion to $4.4 trillion in annual economic benefit, and integrating gen AI and other technologies into business activities could enable the automation of up to 70% of work hours in today's global economy.

• McKinsey will develop lighthouse solutions built on Copilot Studio to advance clients through the next phase of gen AI, leveraging its experience with over 7,000 tech experts, engineers, and data scientists who partner with 500+ clients annually.

🚨 Survey reveals 78% of Americans expect AI abuses to impact the 2024 presidential election. Read More

• 73% believe AI will be used to manipulate social media, 70% think it will generate fake information and deepfakes, and 62% say it will be used to convince certain voters not to vote.

• 93% of Americans believe candidates who maliciously alter or fake media should face punishment, with 46% saying they should be removed from office and 36% advocating for criminal prosecution.

• 69% of respondents aren't confident most voters can detect faked media, and over half doubt their own ability to detect altered or faked audio, video, and photos.

😨 Study reveals voice cloning of major political figures is still easily accessible on AI platforms, raising concerns for the 2024 election. Read More

• The Center for Countering Digital Hate found that 5 out of 6 AI-powered voice cloning services complied with 193 out of 240 requests to clone voices of political figures and generate false statements.

• Only ElevenLabs blocked the creation of voice clones for public figures in 25 out of 40 cases, while Speechify and PlayHT blocked none of the requests.

• Invideo AI not only failed to block any recordings but even generated an improved script for a fake President Biden warning of bomb threats at polling stations, incorporating his characteristic speech patterns, despite prohibiting misleading content.

Have any feedback? Send us an email