Catch up on select AI news and developments from the past week or so (in no particular order):

Anthropic wins major court ruling supporting fair use in AI training. A US federal judge ruled that Anthropic's use of books to train its Claude model qualifies as "fair use," offering a critical legal precedent favoring AI companies. The court emphasized that using books for training was "exceedingly transformative," though storing copyrighted works in a central library was not permitted. This ruling may influence pending lawsuits against OpenAI, Meta, and others, potentially easing copyright concerns in AI development. Importance for marketers: If upheld, this decision enables AI vendors to keep training models on publicly available content, preserving the pace of innovation that fuels marketing tools.

Meta and OpenAI both win fair use rulings in copyright lawsuits. Meta prevailed in a lawsuit brought by 13 authors, including Sarah Silverman, with a judge ruling their arguments failed to show market harm. The ruling stops short of validating all AI training practices but sets a precedent when creators can't prove economic damage. This follows a similar ruling favoring Anthropic earlier in the week. Importance for marketers: These dual wins reduce near-term legal risks for the AI tools marketers rely on, encouraging more aggressive product integrations and content capabilities.

Salesforce upgrades Agentforce platform with observability, open standards, and global model options. Salesforce's Agentforce 3 release offers real-time monitoring, multi-agent coordination, and support for open Model Context Protocol (MCP), enabling secure integration with external tools. PepsiCo is among early adopters, while support for Anthropic and Google Gemini models expands options for regulated industries. Importance for marketers: The enhancements will streamline enterprise-scale AI deployments in marketing and CX operations, helping marketers ensure consistent agent performance and data governance.

Scale AI exposed sensitive client data via public Google Docs, raising security concerns. Business Insider found that Scale AI routinely used public Google Docs to manage projects with clients like Meta, Google, and xAI, leaving thousands of contractor emails, payment disputes, and even confidential training docs accessible to anyone with the link. Importance for marketers: The exposure underscores the importance of data governance when working with AI partners. Brands using AI training or annotation services should review vendor security protocols closely.

Gartner predicts 40% of agentic AI projects will be canceled by 2027. Citing rising costs and unclear ROI, Gartner forecasts that nearly half of all agentic AI projects will fail by 2027. Most current initiatives are still experimental, and many vendors are "agent washing" existing tools without true agentic capabilities. Importance for marketers: While AI agents are trending, marketers should vet claims carefully and invest only where agentic tools clearly enhance workflows or outcomes.

Creatio launches AI-native CRM update, challenges Salesforce's AI add-ons. Creatio's 8.3 Twin release embeds AI deeply into its CRM system, introducing conversational interfaces, no-code agent building, and persistent multi-channel interaction. The platform is priced at $25/month per user and includes AI without upsells. Importance for marketers: This reflects a growing shift toward embedded, affordable AI in marketing platforms—reducing complexity and encouraging adoption in midmarket teams.

Perplexity's valuation hits $14B despite criticism, amid Big Tech acquisition interest. Perplexity AI has raised $500 million at a $14 billion valuation, with reported acquisition interest from Apple and Meta. Critics call it a glorified API aggregator, but its slick interface and ad-free search appeal are winning users. Importance for marketers: If widely adopted, Perplexity could disrupt how consumers and professionals discover content—prompting shifts in SEO and content visibility strategies.

OpenAI considers adding ads to ChatGPT to offset operational costs. CEO Sam Altman acknowledged that ads might be introduced in future versions of ChatGPT, citing growing infrastructure costs. He emphasized a cautious rollout that doesn't compromise trust. Importance for marketers: Ad placements within ChatGPT could create new opportunities—and challenges—for brand visibility and message control in AI-generated content.

Elon Musk plans to "rewrite human knowledge" and retrain Grok on it. Musk said Grok 3.5 or 4 will be trained on a revised "corpus of human knowledge," stripping what he calls "garbage" from current data. Critics compared the plan to rewriting history. Importance for marketers: If Grok becomes a widely used AI platform, its unique training data could skew information access, especially on controversial or politicized topics.

Samsung weighs subscription tiers for Galaxy AI features in 2026 rollout. Samsung will introduce a paid subscription model for its Galaxy AI features, likely beginning in 2026. Not all features will require payment; some popular ones may remain free. Importance for marketers: Samsung's model signals a broader trend toward monetizing AI features in consumer tech—affecting how marketers consider reach, access, and platform capabilities.

OpenAI warns of China's Zhipu AI gaining international contracts. OpenAI reported that China's Zhipu AI is making progress in emerging markets with government and enterprise deployments. The startup, backed by $1.4B in state funds, is part of a broader effort to assert Chinese influence via sovereign AI. Importance for marketers: Rising Chinese AI alternatives may impact global market access, partnerships, and regulatory alignment, particularly for international campaigns and tech stack decisions.

US Senate debates AI regulation moratorium tied to broadband funding. A controversial bill would block states from regulating AI in exchange for broadband infrastructure funds. Critics, including the Teamsters, say it's a giveaway to Big Tech, while supporters argue it avoids a patchwork of state laws. Importance for marketers: If passed, the bill could accelerate national AI rollouts by minimizing compliance complexity—but also stoke consumer distrust in under-regulated environments.