Catch up on select AI news and developments from the past week or so (in no particular order):
Apple eyes AI-powered search alternatives, threatening Google's dominance. Apple is reportedly exploring AI search providers like OpenAI and Perplexity for Safari, signaling a potential break from its $20B/year default search deal with Google. Apple execs confirmed Safari searches declined recently as users shifted to AI tools. Importance for marketers: A shift in default search engines could fragment search behavior, requiring marketers to broaden optimization efforts beyond Google to new AI-native search platforms.
OpenAI announces global Stargate expansion with new country partnerships. OpenAI has launched "OpenAI for Countries," offering nations help to build AI infrastructure aligned with democratic values. The initiative includes localized ChatGPT, national data centers, startup funding, and AI safety standards. Importance for marketers: This move positions OpenAI as a global infrastructure partner, potentially enabling localized, compliant AI capabilities in key growth markets that marketing teams may want to engage.
US Copyright Office report on AI sparks political backlash. A balanced AI copyright report by the US Copyright Office was followed by the sudden firing of its head by the Trump administration, prompting legal and political disputes. The report had called for clearer AI content licensing mechanisms. Importance for marketers: Copyright rules for AI remain unsettled. Content creators and marketers should watch closely to ensure future campaigns remain compliant with evolving IP standards.
Proposed federal ban on state AI laws sparks alarm. A provision in a US House bill would prohibit state-level AI regulation for a decade, drawing criticism for undermining protections like data privacy and AI transparency laws already passed in over 40 states. Importance for marketers: A freeze on state AI laws could create short-term clarity but long-term risk, especially if federal regulations remain vague. Marketers must stay alert to shifts in compliance landscapes.
Meta faces EU legal threat over AI training with user data. Privacy group NOYB has warned Meta it may seek an injunction over its plan to use Europeans' data for AI training starting May 27. Meta cites "legitimate interest" under EU law, but critics argue this violates privacy rulings. Potential damages could reach billions. Importance for marketers: Legal battles over data usage in AI training could limit how platforms evolve, affecting future targeting, personalization, and compliance strategies for marketers operating in the EU.
Meta pulls deepfake ad after Jamie Lee Curtis protests AI misuse. Jamie Lee Curtis publicly condemned Meta for hosting an unauthorized AI ad that used altered footage of her to promote a product. Meta removed the ad after public outcry, but the incident highlights growing concerns around deepfakes and AI manipulation. Other celebrities have also faced unauthorized AI replications. Importance for marketers: This underscores the need for ethical guardrails in AI-driven content, especially around influencer likeness and brand representation.
Newly elected Pope warns AI poses threats to dignity and labor. Pope Leo XIV, the first American pontiff, warned that AI threatens human dignity and labor in his inaugural speech. He pledged the Church's guidance on AI ethics, referencing historical papal engagement with social justice during industrial upheaval. The Vatican is already digitizing archives using AI. Importance for marketers: Ethical AI use is under greater global scrutiny. Marketers using AI must consider reputational risks and the rising importance of AI ethics in public discourse.
Le Monde signs AI content partnership with Perplexity. French newspaper Le Monde has partnered with US AI firm Perplexity, granting the startup access to its content while receiving tech support for new AI products. Perplexity now has deals with over a dozen publishers and shares ad revenue with them. Importance for marketers: As more media firms license content to AI startups, the quality of AI-generated responses may improve—impacting both content strategy and search discoverability.
AI can spontaneously develop human-like social behaviors, study finds. A study from City St George's and IT University of Copenhagen reveals that large language models (LLMs) can develop shared social conventions without external input. In experiments involving dozens of AI agents, naming conventions emerged naturally, mimicking human social behavior. The agents coordinated via pairwise interactions, leading to collective norms and even observable biases. Researchers say this indicates LLMs function more like societies than isolated tools. Importance for marketers: Understanding emergent AI behavior can help marketers anticipate how multi-agent AI systems may evolve in customer-facing tools or autonomous content generation environments.
Meta delays Behemoth AI model over capability concerns. Meta has postponed the release of its flagship AI model "Behemoth," originally scheduled for April. Internal struggles to demonstrate significant improvements over Llama 4 versions reportedly caused delays, raising doubts about the model's readiness. A fall release is now targeted. Meta previewed the model as a major leap, but staff reportedly questioned whether it warranted a public launch. Importance for marketers: Delays in flagship AI model releases suggest potential shifts in Meta's AI strategy, which could impact upcoming tools for content creation, ad targeting, or campaign automation.
Google tests AI Mode in Search, may replace 'I'm Feeling Lucky'. Google is experimenting with an AI Mode button that might replace its iconic "I'm Feeling Lucky" feature. Visible to select users in Google Labs, AI Mode integrates a chatbot interface directly into search. Its varied designs include colorful animations, aiming to stand out. Google says this is part of a limited experiment. Importance for marketers: If adopted broadly, AI Mode may shift search behavior away from traditional SEO pathways, prompting marketers to rethink how they optimize for AI-driven query responses.
TikTok unveils AI Alive filter for animating static photos. TikTok's new AI Alive filter turns static images into short video clips via text prompts. Built-in safeguards and C2PA metadata aim to limit misuse. Users access it through TikTok's Story Camera. Importance for marketers: Tools like AI Alive enable rapid visual content creation for social storytelling, giving brands more interactive options to repurpose existing image assets.
Audible to expand audiobook library using AI voices and translation tools. Audible announced new AI production tools that let publishers generate audiobooks using more than 100 AI voices in multiple languages. Two production modes—end-to-end managed service and self-service—offer flexibility. A translation tool for converting audiobooks between languages is also launching. Narrators who trained AI replicas can review and edit final outputs. Importance for marketers: AI-generated voice tools open up opportunities for global content repurposing and faster audiobook production, benefiting content marketers with audio-first strategies.
Microsoft tests "Hey Copilot!" voice activation on Windows 11. Microsoft is rolling out a wake-word feature for its Copilot assistant, letting users activate it with "Hey Copilot!" The feature uses on-device detection for privacy and works offline, though AI functions still need internet access. Only Windows Insiders with version 1.25051.10.0 or higher can test it currently. Importance for marketers: Increased voice integration with Windows signals broader voice-AI adoption, potentially opening new channels for voice-activated advertising and productivity tools.
Critics say OpenAI's revised structure still poses mission risks. Critics of OpenAI's restructuring say its plan to convert its for-profit arm into a Public Benefit Corporation reduces nonprofit oversight and prioritizes capital-raising over public good. A letter to state attorneys-general argues that safeguards for ethical AI development are still insufficient. Importance for marketers: Governance changes at major AI firms may affect public trust, regulatory scrutiny, and enterprise adoption decisions for AI-powered marketing tools.
Zalando uses AI to accelerate content production and cut costs. Zalando is using generative AI to reduce content production timelines from weeks to days, slashing costs by 90%. It also created "digital twins" of models to streamline product page imagery. Around 70% of campaign images last quarter were AI-generated. Importance for marketers: AI-generated imagery allows for rapid campaign adaptation to trends, creating new efficiencies for marketing teams with tight timelines or limited budgets.
Musk's Grok chatbot promoted conspiracy theory in error. Elon Musk's AI chatbot Grok began referencing "white genocide" in South Africa in unrelated contexts due to faulty prompt conditioning. The issue has since been corrected, but it raised alarm due to the promotion of a known conspiracy theory. Importance for marketers: Incidents like this reinforce the reputational and legal risks of deploying chatbots without rigorous content safeguards.
Alibaba claims 88% reduction in AI training cost with ZeroSearch. Alibaba's new method for AI search training, ZeroSearch, uses simulated documents rather than real search results. The company claims it cuts training costs by 88%, though GPU requirements may impact energy efficiency. Importance for marketers: Cheaper AI training could drive faster innovation and wider AI adoption, influencing how future search and recommendation engines serve marketing content.
Google powers Honor phones with image-to-video AI generator. Google's Veo 2 AI video model is powering an image-to-video tool in new Honor phones, allowing users to animate photos. It's limited to 10 generations per day for now and offers no text prompt input. Importance for marketers: The rise of image-to-video generation via mobile apps hints at new creative formats for brand storytelling and user engagement.
Stability AI releases fast, royalty-free audio model for mobile. Stability AI launched Stable Audio Open Small, a music and sound effects model optimized for smartphones. It uses royalty-free training data and runs locally, though its license restricts use for businesses over $1M in revenue. Importance for marketers: The model enables real-time sound generation on mobile, opening new creative options for campaign sound design without the IP risks seen in competitors' tools.
OpenAI releases GPT-4.1 in ChatGPT with faster coding support. OpenAI's GPT-4.1 and 4.1 mini are now live in ChatGPT. The models, optimized for coding and instruction following, are available to paid and free users, replacing GPT-4.0 mini. OpenAI also launched a new Safety Evaluations Hub. Importance for marketers: GPT-4.1 offers improved speed and instruction following, enhancing marketer workflows that depend on AI for copywriting, analytics, or coding automation.
You can find the previous issue of AI Update here.
Editor's note: GPT-4o was used to help compile this week's AI Update.
