Welcome to The Week in Data, MarTech and AI, a weekly roundup from Quad Insights that sums up the latest news surrounding the technology–driven transformation of marketing.
Edelman Trust Barometer warns of “crisis of grievance”
The just-released 2025 edition of the Edelman Trust Barometer highlights a “crisis of grievance” worldwide, marked by declining faith in institutions such as media, government and business. According to the yearly study, 61% of people worldwide hold grievances against government, business and/or the rich, and 40% see “hostile activism” as acceptable.
Based on surveys of more than 33,000 adults across 28 countries, the barometer provides an updated set of insights about trust, including many that hold implications for marketers.
Here are some of the most striking findings from the participants:
- 47% globally say they trust “owned media.” (By comparison, 42% trust social media, while 63% trust search engines).
- 68% globally worry that business leaders lie to them (69% for government leaders, 70% for journalists).
- 55% in the United States trust business (vs. 50% for non-profits, 41% for government, 42% for news media).
- 66% globally trust the consumer-packaged goods industry (vs. 64% for financial services, 73% for healthcare, 65% for professional services).
- 53% globally trust CEOs “to do what is right” (vs. 77% for scientists, 61% for “my neighbors,” 52% for journalists).
- 58% globally fear their jobs are threatened by automation.
- 63% globally worry about discrimination, prejudice or bias.
More on trust in business
- “2025 Edelman Trust Barometer Launch Event” (YouTube)
- “Insurance customers mixed on whether to trust AI” (Business Insurance)
- “How to Harness the Power of Influencers to Build Trust and Engagement” (Entrepreneur)
AI startup focused on customer experience in retail delivery gets venture investment
An Israeli company focused on applying generative AI to improve the customer experience in order fulfillment has landed a $14 million Series A investment from Susquehanna Growth Equity, Kyt Dotson of SiliconAngle reports. Package.ai has developed a platform that it says offers AI-driven “centralized and automated customer interaction, data and workflows” for an engaging “last-mile” experience in the delivery of purchased goods. The Package.ai platform targets home furnishing retailers, with clients that include nationwide chain Ashley DSG and regional retailers such as Big Sandy Superstore in the Midwest and Bob Mills Furniture in the Southeast.
More on AI in retail
- “Redefining Retail: How AI Is Transforming the Shopping Experience” (Analytics Insight)
- “What to expect in retail in the age of AI shopping assistants” (Retail Brew)
- “AI to become core business driver within next year, claims IBM study” (Retail Customer Experience)
Nielsen gets Media Rating Council sign-off on new TV audience measurement
The Media Rating Council just accredited Nielsen’s Big Data + Panel national audience measurement technology, a new method that combines viewership data from cable, satellite and smart TVs with Nielsen’s traditional consumer panel survey, Brian Steinberg of Variety reports. Nielsen is facing increasing competition from audience measurement upstarts, Steinberg notes; he writes that the “MRC decision appears to show the measurement giant taking some of the strides that TV networks demanded” in terms of modernization.
More on audience measurement
- “Kantar Media sold for $1B amid measurement landscape upheaval” (Marketing Dive)
- “Hy-Vee integrates audience targeting into retail media network” (Chain Store Age)
And finally… AI models don’t know (much) about history
Large language models such as ChatGPT have shown the ability to pass tests such as the bar exam and the SAT. Their knowledge of history, though, leaves something to be desired, according to a new study. Researchers at institutions including the University of Oxford conducted a study in which they challenged the ChatGPT-4, Gemini and Llama chatbots with detailed questions from the Seshat Global History Databank, described as “the most current and comprehensive body of knowledge about human history.” The models got the right answers only between 33% and 46% of the time, Joel Scanlon of AZoAI reports. LLMs are “great for basic facts, but when it comes to more nuanced, PhD-level historical inquiry, they’re not yet up to the task,” study co-author Maria del Rio-Chanona of University College London tells Charles Rollet of TechCrunch.
More on AI capabilities
- “When A.I. Passes This Test, Look Out” (The New York Times)
- “Not enough hospitals are testing their predictive AI models for accuracy, bias, study finds” (Fierce Healthcare)
- “How should we test AI for human-level intelligence? OpenAI’s o3 electrifies quest” (Nature)
More general AI news
- “OpenAI, SoftBank Building AI Data Centers in $500B Stargate Deal” (PC Magazine)
- “Apple Intelligence is enabled by default in iOS 18.3” (The Verge)
- ICYMI: “AI continues to aid marketers’ quest for authenticity: Report” (MarTech)
- ICYMI: “AI agents are beginning to infiltrate the advertising world” (Marketing Brew)
