AI News from Fast Company AI

Latest coverage from Fast Company AI, summarized and scored for signal.

• The data center backlash is sending AI infrastructure to some unexpected places — With 70% of Americans opposing local data center construction, AI infrastructure developers are exploring offshore and orbital alternatives — a signal of how acute the land-use conflict has become.
• Mozilla is bringing AI to Firefox—but only if you want it — Firefox's Smart Window AI mode adds opt-in AI chat with live Exa-powered web search and natural language browsing history recall — with a master kill switch in settings that turns off all AI features for users who want the old Firefox experience.
• AI coding tools unlock small software — AI coding tools are collapsing development costs toward zero, making millions of previously uneconomical small software programs worth building — a shift that restructures who builds software and what gets built.
• Tokenmaxxing is out, valuemaxxing is in — Tesla, Uber, and Meta are capping AI token spending and demanding ROI accountability after months of rewarding raw usage volume — a corporate reckoning that signals the end of the more tokens equals more progress era.
• AI companies are buying used books by the thousands. Some may be destroyed for training — AI companies are bulk-purchasing used books from secondhand bookstores to scan as legally acquired training data, giving booksellers a surprise sales boost while raising questions about whether digitizing purchased physical copies infringes copyright.
• Not every AI-in-education tale is a horror story. How the world’s leading education company makes AI that’s actually useful for students — Pearson, the 180-year-old education giant, is building AI learning tools aimed at improving student outcomes — though the source provides minimal detail about what the products actually do or what outcomes have been measured.
• Your AI is emailing my AI—and nobody’s in charge — With 50% of U.S. workers now using AI to draft workplace messages, an increasing share of professional communication is AI-to-AI exchange — raising real questions about accountability, authenticity, and who actually owns the message.
• At Black Hat, AI is helping security pros find threats—and creating new ones — Black Hat 2026 put AI's dual cybersecurity role in sharp relief: conference network admins are using AI to triage thousands of daily security alerts, while AI-generated apps with poor security hygiene are emerging as a new attack surface for the practitioners hunting them.
• Chatbots argue against election conspiracy theories, then willingly illustrate them — A Brennan Center study found major AI chatbots declined to endorse false election conspiracy theories — but the same companies' image generation tools produced fabricated visual 'evidence' for those exact claims when prompted.
• Do you know what your AI is doing right now? — Enterprise AI adoption is stalling because most companies cannot answer basic questions about which AI agents are running in their infrastructure, what data they are accessing, or what permissions they hold.
• AI crawlers from Meta and Alibaba almost destroyed a volunteer-run LGBT history archive — Aggressive AI web crawlers attributed to Meta and Alibaba nearly shut down a volunteer-run LGBT history archive by driving up hosting costs enough that the site founder spent nights building bot defenses — highlighting collateral damage of AI training data collection on small public-interest sites.
• The world’s largest ‘biological datacenter’ could help make animal testing obsolete — Vivodyne is building what it calls the world's largest biological datacenter — running AI-orchestrated drug experiments on lab-grown human tissue at scale — aiming to replace animal testing with human-tissue models that produce more predictive drug candidates.
• Target names its first chief AI officer in a bid to remake the shopping experience (exclusive) — Target has appointed former Lowe's executive Chandhu Nair as its first-ever chief AI officer, tasking him with applying AI across employee tools, inventory management, and the online customer shopping experience.
• Mark Zuckerberg lays out Meta’s AI vision in a 6,500-word essay. 6 things to know — In a 6,500-word essay, Meta CEO Mark Zuckerberg makes the case for open-weight AI as a public good, argues heavy regulation primarily benefits incumbents, and pledges $1 billion to communities hosting Meta data centers.
• Reimagined work will create AI transformation — Fast Company argues organizations should treat AI as an opportunity to redesign how work is structured rather than deploying AI tools into existing workflows unchanged — a framing shift that is easier to prescribe than to execute but increasingly backed by enterprise ROI data.
• Social fandom beats big movie budgets — Iron Lung's grassroots social buzz outperformed Supergirl's $115M marketing spend at the box office, prompting entertainment companies to mine footage archives and fan communities as cost-effective competitive advantages — with AI tools increasingly cited as enablers of content repurposing and audience analysis.
• ChatGPT’s traffic surge is good news for brands. For publishers, it’s complicated — As ChatGPT drives more referral traffic, brands gain a new AI-powered discovery channel while publishers face a structural paradox: their content is the credibility layer that makes OpenAI's emerging ad business work, but the revenue model for that contribution remains undefined.
• Some adults use AI for financial advice, but trust is still low, Gallup poll finds — A Gallup poll finds some US adults use AI for financial advice, but overall trust remains low — revealing a gap between AI capability in personal finance and consumer confidence that fintech products have yet to bridge.
• Google’s AI leadership comes apart in a single morning — Google's AI leadership structure cracked in a single morning: Jeff Dean is departing after 27 years to start his own company, Demis Hassabis is handing off DeepMind day-to-day leadership, and Gemini 3.5 Pro still has no release date.
• The next ad market may be built for machines — Publishers are beginning to sell ad space directly to AI agents rather than human readers, with Time magazine piloting machine-targeted ads that raise questions about transparency and AI influence on content recommendations.
• The AI ‘tokenmaxxing’ corporate fad is fading as workplaces look to cut costs — Fast Company reports that 'tokenmaxxing'—the Silicon Valley trend of treating high AI token consumption as a proxy for employee performance—is fading as enterprises shift focus to controlling AI costs rather than maximizing AI output volume.
• OpenAI tells ChatGPT to stop impersonating famous authors — OpenAI has directed ChatGPT to stop impersonating famous authors, a concrete policy response to mounting copyright pressure from writers and publishers as AI companies race to pre-empt legal action over voice and style mimicry.
• Anthropic releases Claude Opus 5 for both AI coding and general office work — Anthropic has launched Claude Opus 5, its flagship model, targeting both AI-assisted coding and general office productivity — the company describes it as capable of working through complex multi-step problems with substantially less back-and-forth user interaction.
• Forget layoffs: This quiet workplace tactic is shucking headcounts without anyone noticing — Smaller companies are quietly reducing headcount without formal layoffs by simply not replacing workers who leave — routing the departed employee's workload to remaining staff and AI tools, according to a new survey reported by Fast Company.
• Can the U.S. government legally gatekeep global access to AI models? — For the first time, the U.S. government used export controls to disable Claude Fable 5 and Mythos 5 for foreign users — blocking international access to specific AI services and raising urgent legal questions about Washington's authority to restrict global AI access under existing trade law.