Meta will cut 10% of its workforce - approximately 8,000 employees - next month, a move the company said is designed to offset its aggressive spending on artificial intelligence. The layoffs follow a broader pattern across Big Tech, where combined AI infrastructure spending plans for 2026 have now swelled to $725 billion, a 77% jump from the previous year's record capital expenditures.
The internal memo, confirmed by a Meta spokesperson, told staff the cuts were necessary to "run the company more efficiently and [...] offset the other investments we're making." The decision comes after CEO Mark Zuckerberg said Meta would increase its spending to as much as $135 billion this year. Meta's share price has been volatile since the announcement, reflecting investor nervousness around the massive increase in spending before AI's promises are fully realized.
AI safety pledges meet practical hurdles
Employees at OpenAI and Anthropic are questioning how their leaders' recent calls to slow AI development will work in practice. According to the Financial Times, staff were blindsided by proposals that include embedding independent evaluators inside AI labs to scrutinize models and safety practices. Workers are now grappling with questions around how much access outsiders will receive and how intellectual property will be protected.
The uncertainty highlights a growing gap between the consensus that safety measures must keep pace with advancing models and the practicalities of implementing those commitments. It also raises questions over whether the rare show of agreement between AI leaders will translate into meaningful change around model development and release.
AI reshapes e-commerce and brand discovery
Google unveiled Universal Cart at its I/O 2026 conference, an AI-powered shopping hub that lets users add products to a single cross-platform cart across Search, Gemini, YouTube, and Gmail. AI models track pricing, surface loyalty rewards, and flag compatibility issues. The checkout features will soon be available across merchants like Nike, Sephora, Target, and Walmart, with the brand remaining the merchant of record.
Estée Lauder Companies is also recalibrating its digital strategy, partnering with AI marketing platform Profound to track how its beauty brands appear on generative AI platforms like ChatGPT and Gemini. "The way people discover beauty is being rewritten, and we intend to shape that shift rather than react to it," said ELC global chief digital and marketing officer Aude Gandon. The partnership will use agentic tools to optimize product pages, blogs, and social media so that information is easier for large language models to comprehend.
Meanwhile, AI shopping startup Daydream is integrating deeper into the iPhone with features built on Apple's new iOS 27 tools. Users can now find shoppable matches for clothing in photos saved to their camera rolls or ask Siri to search Daydream's catalog using conversational prompts without opening the app. Daydream has also launched a B2B arm, Powered by Daydream, a white-label platform allowing fashion brands to embed its AI search directly into their e-commerce sites. Launch partners include Staud, Alice + Olivia, and Cult Mia.
Legal battles over content, copyright, and children
A group of major publishers - including Hachette, Macmillan, McGraw Hill, and Elsevier - has sued Meta, alleging the company illegally used millions of pirated books and academic texts to train its Llama AI models without permission. The lawsuit marks an escalation in the battle between creative industries and AI developers, moving from individual creators to coordinated action from large media and publishing groups.
In a separate case, a landmark child privacy lawsuit against Meta has kicked off in the US. Twenty-nine states claim Meta violated multiple federal and state laws protecting children. Beyond potential billions in payouts, the states are demanding changes to Instagram and Facebook, such as removing like counts and the "infinite scroll." The UK has also moved to ban social media for under-16s, effective next year, modeled on restrictions set by the Australian government.
Amazon faces its own legal challenge, with a class action lawsuit brought on behalf of millions of Twitch streamers who allege the platform's parent company used their content to train AI without permission or compensation. According to the BBC, 215 million hours of content was generated via the platform in the first few months of 2026 alone.
Why this matters for creatives, marketers, and writers
The infrastructure of how brands reach consumers is being rebuilt. Google's Universal Cart and the push toward agentic commerce mean that AI-mediated recommendation systems - not just keywords or storefronts - will increasingly shape purchasing decisions. For marketers and PR professionals, this requires a shift toward generative engine optimization and a clear strategy for how brands appear in AI assistant responses. Professionals looking to build these skills can explore AI Training for Brand Managers and AI for PR Courses.
For writers and creative professionals, the copyright lawsuits against Meta signal that the question of whether AI companies can train models on copyrighted material under "fair use" is far from settled. The outcome will help determine how creative content is valued online. Meanwhile, OpenAI's release of ChatGPT Images 2.0 - with its ability to parse visual references posted to the internet as recently as December 2025 - will intensify concerns around AI's immediate reproduction of creative IP. Cate Blanchett's new non-profit, RSL Media, is developing a "human consent standard" for artists to formally declare permissions for how their identity and creative works can be used by AI models.
Your membership also unlocks:
