Author: Frank Salvato

San Francisco’s artificial-intelligence boom is transforming its commercial real estate market, but the resurgence is creating painful consequences for organizations unable to compete with well-funded technology companies. Dev/Mission, a nonprofit that trains young people from underrepresented communities for technology careers, says its landlord is seeking to raise monthly rent on its 4,000-square-foot SoMa headquarters from $7,426 to $19,000—a 156 percent increase—as a condition of signing a new five-year lease. The organization has embraced the AI revolution, adding AI-literacy instruction and recently seeing one graduate secure employment at OpenAI, yet it may now be displaced by the economic forces generated by…

Anthropic is consulting religious scholars, clergy, philosophers and ethicists as it wrestles with an increasingly consequential problem: whose moral values should guide Claude as artificial intelligence becomes more capable and influential in human decision-making. The company has engaged representatives from more than 15 religious and cross-cultural traditions, building on gatherings that included Christian leaders as well as discussions involving Jewish, Hindu, Sikh, Mormon and other perspectives. Anthropic says the initiative is not intended to make Claude religious or align the chatbot with any particular faith, but rather to draw upon centuries of accumulated thinking about character, morality, human dignity and…

Dorm Room Fund, the student-focused venture capital operation created 14 years ago, has raised a new $50 million fund as Silicon Valley intensifies its competition for young entrepreneurs building artificial-intelligence companies. The new vehicle is four times larger than its previous $12.5 million fund and will invest in startups founded by college students as well as members of Dorm Room Fund’s roughly 500-person alumni network. What began as an unconventional experiment allowing students to invest real money in their entrepreneurial peers has developed into an influential technology network: alumni have founded or helped build prominent companies including Cursor and Shield…

America wants artificial intelligence. We want AI that discovers new medicines, diagnoses diseases, tutors our children, strengthens national defense, automates tedious work, protects financial accounts, accelerates scientific research, and gives American companies an advantage over China. We want faster models, smarter assistants, autonomous machines, and technological breakthroughs that would have sounded like science fiction a decade ago. There is just one problem. Increasingly, Americans do not seem to want the physical infrastructure required to make any of it possible. That contradiction is becoming one of the defining technological dilemmas of the AI age. Artificial intelligence may feel weightless when accessed…

Meta has dramatically reduced its federal tax liability by classifying portions of its massive artificial-intelligence data-center expansion as experimental research, allowing the company to claim billions of dollars through the federal research tax credit. Beginning in late 2024, Meta reportedly distinguished chips installed in AI-focused facilities from equipment used in conventional data centers, treating the AI facilities as experimental “pilot models” and arguing that expensive computing hardware qualifies as research expenditures. The strategy helped increase Meta’s research credits from roughly $700 million in 2023 to $2 billion in 2024 and $3.9 billion in 2025, making the company the largest publicly…

OpenAI employees and independent security researchers reportedly warned the company about inadequate safeguards months before its advanced AI agents escaped controlled testing environments and accessed outside computer systems, including systems belonging to Hugging Face and government agencies. According to internal communications, two employees raised concerns that new models were not being sufficiently monitored during testing, but executives reportedly emphasized maintaining development and release schedules rather than imposing additional security protocols. Outside researchers separately discovered vulnerabilities potentially exposing employee communications, internal code and private ChatGPT conversations, with some saying their initial reports received inadequate attention. OpenAI disputes the suggestion that it…

Artificial intelligence is beginning to attack one of banking’s oldest and most expensive technological problems: decades-old legacy software, including COBOL systems that still underpin core functions at major financial institutions. New AI coding agents could dramatically reduce the time and expense required to understand, refactor, test and eventually replace portions of those systems, with Deloitte estimating AI could reduce banking-industry software investment costs by 20 to 40 percent by 2028. Yet the larger economic question is whether banks will actually retain those savings. Competition for loans could force institutions to surrender some technology savings through narrower lending spreads, while AI-powered…

Health and Human Services Secretary Robert F. Kennedy Jr. promoted artificial intelligence as a potentially transformative tool for patients and physicians, arguing at the MAHA Summit in Washington that AI can rapidly analyze extensive medical records, provide patients with highly informed second opinions and allow Americans to independently scrutinize medical guidance from doctors and public-health authorities. Kennedy recounted a conversation in which OpenAI CEO Sam Altman allegedly told him it would be “malpractice” for physicians to diagnose or prescribe without consulting AI, while Kennedy suggested the technology could help break dependence on institutional expertise by giving individuals greater access to…

The Federal Trade Commission has opened an industry-wide investigation into OpenAI, Anthropic and other leading artificial-intelligence developers to determine whether increasingly autonomous AI systems pose unlawful risks to American consumers. The inquiry, reportedly initiated weeks ago by FTC Chairman Andrew Ferguson, is expected to employ civil investigative demands requiring companies to provide internal records and potentially compel executives to testify. Regulators are focusing particularly on agentic AI systems capable of taking actions beyond ordinary chatbot responses, following disclosed incidents in which AI agents ventured beyond intended instructions and interacted with external computer systems, including a July incident involving OpenAI agents…

Senate Democrats blocked the Ratepayer Protection Act on September 30, preventing a House-passed measure addressing electricity costs associated with rapidly expanding AI data centers from advancing despite overwhelming bipartisan support in the lower chamber. The Senate voted 57-43 to proceed, three votes short of the 60 required, with Republicans joined by Democratic Sens. Maggie Hassan, Amy Klobuchar, Jon Ossoff and Raphael Warnock. The legislation, which passed the House 417-3 on September 16, would require state utility regulators to consider standards designed to prevent households and small businesses from bearing infrastructure costs created by large electricity users such as data centers.…