Welcome to the latest edition of the AI Update, where we continue to track the major developments in artificial intelligence and their implications for identity security and cybersecurity more broadly.
The Pope issued an AI-focused encyclical, arguing that AI technology must serve humanity rather than concentrate power. Titled Magnifica Humanitas: On Safeguarding the Human Person in the Time of Artificial Intelligence, the document is Pope Leo XIV’s first encyclical, and it recognizes the emergence of AI as a world-historical transformation, connecting it to Rerum Novarum, Leo XIII’s 1891 encyclical on labor and social upheaval during the industrial age. And it warns that when AI systems appear neutral and objective, they can hide design choices, stereotypes, and ideological bias, while making it harder for affected people to appeal or identify who is responsible.
Anthropic has surpassed arch-rival OpenAI in its valuation, and it’s so close to being a trillion-dollar company. The startup has raised $65 billion in a Series H funding round that values the company at $965 billion post-money, with the round having been led by Altimeter Capital, Dragoneer, Greenoaks, and Sequoia Capital, with Coatue and ICONIQ as co-leads. Anthropic’s valuation has more than doubled since February, when it raised $30 billion at a $380 billion post-money valuation.
China-based ByteDance is developing new AI chips focused on inference – the actual thinking, so to speak, that AI models do, rather than the training of them. The chips will reportedly be modelled on the architecture of those made by Groq, an AI chip startup whose technology Nvidia licensed late last year. The move could be significant within the broader US-China AI race, given how Chinese companies have seen their progress hampered by American sanctions that have restricted their access to Nvidia chips.
China is undergoing its biggest surveillance tech upgrade in years, with an FT report detailing how AI technology is being layered onto the country’s vast camera and public-security infrastructure. Drawing from over a dozen procurement documents and industry interviews, the report shows local government efforts to implement AI-powered tech meant to enable police to detect behaviour patterns and search footage using written prompts. This is building a kind of data interpretability layer onto an already substantial network of surveillance systems including facial recognition cameras.
European Union leaders have prepared a draft “European tech sovereignty strategy” that would shift away from heavy regulation and toward fostering European AI infrastructure to reduce dependence on the US. Among other things, this could reportedly include the introduction of a Cloud and AI Development Act that would encourage faster permitting for data centres, sovereignty risk assessments for technology use, cloud-service classifications based on control and data handling, and a revised semiconductor law. The Cloud and AI Development Act is expected to be unveiled on June 3, after delays caused by internal disagreements.
Illinois lawmakers have passed SB 315, the Artificial Intelligence Safety Measures Act, which would require large AI developers to publish and annually update safety frameworks, issue transparency reports before deploying new frontier models, report critical safety incidents, protect whistleblowers, and undergo annual independent third-party audits. The bill has passed both the Senate and the House, and Governor JB Pritzker says he will sign it, with an implementation date set for January 1, 2027. Anthropic and OpenAI are reportedly okay with the bill, while groups such as Chamber of Progress oppose it, arguing that it could expose proprietary systems without clear regulatory standards.
The chatbot’s take: We asked for a little bit of an encyclical explainer for the non-Catholics among us.
–
By Alex Perala
