79% of Young People Use AI for Shopping, Mastercard Finds

Key Takeaways

• <ul><li>79 percent of young people and 70 percent of parents have used AI for product research in the past year, per Mastercard.</li><li>75 percent of teenagers use AI to research products, and 45 percent would trust AI to make purchases for them.</li><li>45 percent of teens worry about financial data safety when using AI for shopping.</li><li>Fraudsters now use AI for multichannel "3D phishing" attacks combining text, voice, and deepfakes that change tactics hourly.</li><li>The UK and Canada are renewing their AI security partnership to develop shared standards and cooperate on frontier AI risks.</li><li>Massachusetts proposes requiring independent evaluators to test AI models for catastrophic risk at least three times a year.</li><li>Google, Anthropic, OpenAI, and Meta are lobbying against Massachusetts' proposed AI safety bill.</li><li>Axe Compute partnered with Penguin for AI infrastructure sites covering over $3.45 billion in contracted projects.</li><li>DeepLumen's Agentic Page is installed by nearly 2,000 Shopify merchants to track AI-driven referral sales, including via Meta's Muse.</li><li>CoverSynth and CRIO's AI eSource Builder improve AI evaluation and clinical research workflows, respectively, without heavy human labeling.</li></ul>

Mastercard Report: 79 Percent of Young People Use AI for Product Research

Mastercard found that 79 percent of young people and 70 percent of parents have used AI to research products in the past year. Many use AI at least once a month to compare options or plan purchases. Most young people would be willing to let an AI assistant manage shopping for them. The report says users must keep control by setting spending limits and approval rules. Retailers will need to provide clear, verifiable information that AI agents can trust.

Teenagers Turn to AI Shopping as Payment Security Concerns Grow

Mastercard's study shows that 75 percent of teenagers use AI to research products before buying. Sixty percent use AI to compare prices, and 45 percent would trust AI to make purchases for them. However, 45 percent also worry about the safety of their financial information when using AI. The findings call for stronger security measures as AI becomes a bigger part of teen shopping habits.

UK and Canada Renew Partnership on AI Security

The United Kingdom and Canada are renewing their AI security partnership, building on agreements from 2024 and 2025. The collaboration combines Canada's AI research with the UK's testing and evaluation expertise. Both countries will focus on developing shared security standards, sharing sensitive information, and expanding technical cooperation on frontier AI risks. The UK also plans to lead AI safety discussions during its G20 Presidency next year.

AI Is Changing Bank Phishing and Payment Fraud Attacks

Fraudsters are now using AI to create convincing phishing messages, fake websites, cloned voices, and deepfakes. Banks face multichannel attacks that blend text, voice, and video in what is called 3D phishing. Defenders need real-time AI systems that monitor risk across the entire customer journey and intervene before money is lost. Banks must move from one-time identity checks to continuous trust and connect data across all channels. Human analysts cannot keep up with AI-generated fraud that changes tactics every hour.

CoverSynth: A New Tool for Building Better AI Evaluation Sets

CoverSynth is a new method for creating evaluation sets that test how well AI retrieval systems perform. It targets coverage across different types of questions and system abilities without relying on expensive human labeling. The approach uses deterministic and model-based checks to verify evidence and guarantee reliable ground truth. Tests across five public corpora show CoverSynth reveals failure modes that human-written sets often miss. On the RepLiQA corpus, CoverSynth-generated queries better distinguish differences between retrieval systems.

Axe Compute partners with Penguin to expand AI infrastructure

Axe Compute has selected Penguin as a strategic partner for multiple AI infrastructure sites. The deal covers over $3.45 billion in contracted customer projects. Penguin will provide ClusterWareAI software and 24/7 network operations center support at GPU cluster sites. The partnership aims to speed up deployments and help Axe Compute take on more Build projects to meet growing demand.

DeepLumen helps merchants sell through AI agent referrals

DeepLumen's Agentic Page product has been installed by nearly 2,000 Shopify merchants to track sales from AI-driven referrals. The tool helps merchants make product information accessible to AI systems like Meta's Muse. Agentic Page is free to install through the Shopify App Store, with DeepLumen charging a commission only on attributable AI orders. The fastest first order came within 24 hours of deployment. DeepLumen is also preparing versions for WooCommerce, Amazon, and independent commerce sites.

Researchers examine whether AI improves weather forecasting

An article explores the question of whether artificial intelligence is making weather forecasts more accurate. The topic addresses growing interest in AI applications for meteorology and prediction. The piece examines how AI tools are being used in the field of weather forecasting.

Massachusetts proposes strict new AI safety regulations

Massachusetts Gov. Maura Healey is pushing for one of the nation's strictest AI safety laws. The proposed legislation would require independent evaluators to test AI models for catastrophic risk at least three times a year. Companies like Google, Anthropic, OpenAI, and Meta are lobbying against the bill. California, Illinois, and New York already have AI safety laws, but Massachusetts would be the first to require outside testing of core software. The bill is still being discussed in the state legislature.

Colossyan leads AI course video generators for corporate training

A 2026 guide ranks the best AI course video generator tools for corporate learning teams. Colossyan ranks first for turning documents into SCORM-ready courses with avatar video and assessments. The platform costs $59 per month for Professional plans on yearly billing. The guide reviewed eight tools based on vendor documentation, pricing, and user reviews. Colossyan was selected because it builds complete courses with assessments rather than just standalone videos.

CRIO Launches AI eSource Builder to Cut Protocol Review Time by Four Hours

CRIO introduced a new AI eSource Builder tool that helps study coordinators create eSource drafts faster. Users upload a protocol and eCRFs, and the tool generates visit schedules, procedure templates, eligibility criteria, and visit logic. Early pilot sites reduced eSource build time by four hours per study. The tool flags unmatched procedures as placeholders and requires structured review before finalizing. Access requires a Publisher site within CRIO, which is free for existing users.

Sources

• Mastercard Report Finds 79% of Young People Use AI to Research Products
• AI shopping gains ground among teenagers as payment security comes into focus
• Great Britain: UK-Canada Renewed AI Security Partnership
• AI Is Rewriting the Playbook for Bank Phishing and Payments Fraud
• CoverSynth: Coverage-Guided Generation of Retrieval Evaluation Sets
• More than $3.45 billion in customer AI infrastructure projects are contracted. Axe Compute expects Penguin to speed deployments.
• As Personal AI Agents Take Off, DeepLumen Helps Merchants Turn AI into a New Sales Channel
• Is artificial intelligence making weather forecasts better?
• Massachusetts Could Set a New Standard for AI Safety
• The Best AI Course Video Generator Options for Corporate Learning Teams in 2026
• CRIO Launches AI eSource Builder to Cut Protocol Review Time by Four Hours