Anthropic is investing to advance next-generation AI systems, expand compute capacity, research mechanistic interpretability and alignment, and accelerate international expansion.
https://www.pymnts.com/news/investment-tracker/2025/anthropic-backed-goodfire-raises-50-million-to-access-ais-internal-thoughts/Geoffrey Hinton, a Nobel Prize-winning computer scientist, has warned that the rapidly developing AI industry poses a significant threat to humanity, citing a 10-20% chance of AI surpassing human control. In an interview on CBS, Hinton compared the situation to having a "cute tiger cub" that may grow into a deadly animal if not properly understood and controlled. He attributes his concerns to the lack of understanding among people about the direction AI development is heading, which he believes will lead to catastrophic consequences if left unchecked.
https://www.techspot.com/news/107706-godfather-ai-geoffrey-hinton-warns-there-10-20.html#commentsOffsetThe US Congress has introduced a bill that critics fear will grant authorities increased censorship power, with some states such as California and Florida already passing laws criminalizing the publication of sexually explicit deepfakes. The Electronic Frontiers Foundation has expressed concern that the legislation will be used to manipulate platforms into removing lawful speech that is disapproved of by those in power, including President Trump, who has stated he would use the law to censor his critics.
https://www.victoriaadvocate.com/news/nation/congress-passes-revenge-porn-ban-sending-it-to-trump/article_d230ac1a-e9ec-5c48-9e52-2caa20bf4bd0.htmlThe Children's Commissioner in the UK has called for an immediate ban on AI apps that enable "deepfake" sexual abuse of children. The issue has affected celebrities and most concerning, children, with a recent scandal forcing a school in Pennsylvania to close temporarily. Dame Rachel de Souza, the Children's Commissioner, warns that girls are avoiding posting images online due to fear of being targeted by these technology, stating there is no positive reason for such apps to exist.
https://petapixel.com/2025/04/28/uk-childrens-commissioner-calls-for-ban-on-ai-nudification-apps/Researchers from the University of Zurich secretly deployed personalized AI arguments on the ChangeMyView subreddit, violating platform rules and university ethics guidelines. The study found that AI bots are highly persuasive and challenging to detect, but its methodology raises ethical concerns, prompting moderators at the ChangeMyView subreddit to speak out against the unauthorized experiment involving AI-generated comments.
https://www.searchenginejournal.com/reddit-mods-accuse-ai-researchers-of-impersonating-sexual-assault-victims/545532/Akamai Technologies has announced Firewall for AI, a new solution to protect large language models and other emerging AI applications from unauthorized queries, adversarial inputs, and data scraping attempts. The company's Senior Vice President Rupesh Chokshi stated that traditional security solutions are not designed to mitigate AI threats, emphasizing the need for AI-driven capabilities. Firewall for AI provides multilayered protection, real-time threat detection, compliance and data protection, flexible deployment options, and proactive risk mitigation. Additionally, Akamai is introducing API LLM Discovery, which automatically discovers and updates security policies for GenAI and LLM API endpoints to prevent unauthorized access and abuse.
https://www.prnewswire.com/news-releases/akamai-firewall-for-ai-enables-secure-ai-applications-with-advanced-threat-protection-302439032.htmlColorado lawmakers are considering changes to a law aimed at preventing automated bias by AI systems, which was introduced in 2024. The proposed legislation would loosen restrictions on the use of AI in employment, finance, housing, and other sectors. Governor Jared Polis supports the bill, while AI developers say it is overly broad.
https://news.bloomberglaw.com/artificial-intelligence/colorado-ai-bill-would-overhaul-states-sweeping-anti-bias-lawThe US Department of Justice is urging more attention to be paid to rising AI-related financial crimes, with former Deputy Attorney General Lisa Monaco warning of a "robust" approach to enforcement. The increasing sophistication of AI-powered scams, including confidence tricks and cyberattacks, poses significant challenges for law enforcement. Financial frauds are becoming easier to commit using AI, such as fake digital personas used in romantic or financial relationships, making it essential for the DOJ to take action to combat these crimes.
https://news.bloomberglaw.com/before-the-bar/rising-ai-financial-crimes-need-more-urgent-attention-from-dojGenpact has released new research on driving AI adoption in the insurance industry, highlighting strategies to overcome customer skepticism. The study aims to reshape perceptions of artificial intelligence in insurance, with a focus on turning uncertainty into confidence among customers.
https://www.prnewswire.com/news-releases/genpact-study-shows-insurance-customers-embrace-ai-when-value-and-benefits-are-clearly-demonstrated-302440905.html