AI is reshaping our world at a dizzying pace. Unlike past tech breakthroughs, government regulation is struggling to keep up, leaving the public vulnerable. From chatbots encouraging harmful behavior to potential misuse in creating weapons, the risks are real. There is no federal body ensuring AI safety before release, and companies are not required to report dangerous incidents. While industry leaders claim to prioritize safety, massive investments and strategic choices often outpace caution. Public trust is low; a recent survey shows 77% of Americans see AI as a potential threat. Independent oversight is urgently needed to protect human rights and guide responsible AI development.
Discover how to engage with AI thoughtfully without compromising your judgment, privacy, or humanity through our free course "AI for the People." This six-week newsletter series guides you in practical, ethical ways to harness AI's potential effectively. Sign up and start transforming your relationship with technology today.
Anthropic has firmly declined the Pentagon’s request to eliminate safety measures from its AI model, Claude, despite the US military’s threat to cancel a $200 million contract and label the company a "supply chain risk." The Department of Defense demanded unrestricted access to Anthropic’s AI technology by Friday, but the company stated it cannot comply in good conscience, emphasizing the importance of maintaining ethical safeguards in its AI deployment.
A recent study reveals that ChatGPT Health, an AI-based health advisory tool, missed recommending hospital visits in over half of critical cases. The platform also frequently failed to recognize signs of suicidal thoughts, raising concerns among experts about potential harm. Launched by OpenAI in January for select users, ChatGPT Health allows individuals to link medical records and wellness apps for tailored advice. Despite being used by over 40 million daily for health inquiries, the AI's limitations highlight serious safety considerations.
A small number of corporations dominate the internet, wielding extensive control over our personal information. Despite their influence, there exist ethical and often European-based options that challenge these giants. Concerns around big tech include social media's negative impacts, misinformation, data exploitation, environmental harm, and tax avoidance. Silicon Valley's cozy relationships with controversial political figures add to the distrust. Moreover, many products from these firms have become less helpful and more exploitative over time, a phenomenon tech commentator Cory Doctorow calls “enshittification.” As consumers, we trade our privacy and data for access to free or low-cost services, ultimately fueling the financial power and monopoly status of these companies.
Detectives credit Palantir's AI-powered tools for playing a crucial role in securing convictions against a criminal gang responsible for an £800,000 theft. The 'Fuck the Police' gang, operating between Luton and Romania, executed over 3,000 fraudulent cash machine withdrawals throughout 2024. Bedfordshire police faced the challenge of sifting through 1.4 terabytes of digital data from 24 seized smartphones to link the suspects across eastern England and Bacau, Romania.
Meta, the parent company of Facebook, has entered into a $60 billion agreement with semiconductor giant AMD to purchase artificial intelligence chips. This strategic investment is seen as a significant bet on AI technology, made despite concerns about the large financial risks associated with AI infrastructure development. Industry analysts note that this move is part of a broader trend among US tech firms, which are projected to invest $660 billion in AI technologies this year.
Shares of major companies like Uber, Mastercard, and American Express dropped following a speculative AI doomsday scenario shared by Citrini Research on Substack. The firm warned of a potential future where autonomous AI systems disrupt the US economy, affecting jobs, financial markets, and mortgages. This cautionary outlook, labeled as a "scenario, not a prediction," has intensified investor concerns about AI's economic impact.
The UK's Information Commissioner's Office (ICO) has issued its largest fine yet, penalizing Reddit £14.5 million for unlawfully processing data from users under the age of 13. The ICO's action follows revelations that the platform previously lacked lawful grounds to handle children's personal information, exposing them to potentially harmful content. Since July, Reddit has implemented age verification measures to restrict mature content access, addressing earlier compliance gaps identified by the regulator.
The conversation about AI potentially taking over human jobs often overlooks a crucial question: who will ensure that people have enough to eat? Despite fears mirroring those from the Industrial Revolution, which ultimately did not displace workers entirely, there is a lack of serious discussion on how society should respond if AI does make many jobs obsolete. Addressing how food distribution and sustenance will be managed in such a future is essential.
