2% of Americans Pay for AI Services, Spending $31 Monthly

Key Takeaways

• Only 2% of Americans pay for AI services, spending about $31 per month on average
• President Trump has denied the need for AI safeguards, but Jay Clayton was named U.S. AI czar to oversee regulation policy
• Cato Networks earned AWS Security Competency in AI Security to protect tools like Amazon Bedrock and Amazon SageMaker
• Meta faces user complaints over AI moderation errors and lack of human support
• Anthropic says its AI agents did not breach Australian government websites
• AI misuse is the top reputational threat for brands, according to a global survey
• The FDA plans new AI guidance for healthcare by 2027
• Salesforce stock remains volatile amid AI-driven uncertainty
• WorldTel launched enterprise AI and carbon market services across five divisions
• Researchers developed tools to predict sales leads and verify code equivalence using AI

Most Americans use AI but few will pay for it

Research from Tech Crunch found that only 2% of Americans were paying for AI services as of May, with an average cost of about $31 per month. The slow growth in paying users is pushing some AI companies to focus on business customers instead. AI companies face challenges because the technology is expensive to run. Both Republicans and Democrats are calling for more regulation of AI. President Trump has denied the need for AI safeguards, but Jay Clayton was named U.S. AI czar to oversee regulation policy.

Most Americans use AI but few will pay for it

Research from Tech Crunch found that only 2% of Americans were paying for AI services as of May, with an average cost of about $31 per month. The slow growth in paying users is pushing some AI companies to focus on business customers instead. AI companies face challenges because the technology is expensive to run. Both Republicans and Democrats are calling for more regulation of AI. President Trump has denied the need for AI safeguards, but Jay Clayton was named U.S. AI czar to oversee regulation policy.

Most Americans use AI but few will pay for it

Research from Tech Crunch found that only 2% of Americans were paying for AI services as of May, with an average cost of about $31 per month. The slow growth in paying users is pushing some AI companies to focus on business customers instead. AI companies face challenges because the technology is expensive to run. Both Republicans and Democrats are calling for more regulation of AI. President Trump has denied the need for AI safeguards, but Jay Clayton was named U.S. AI czar to oversee regulation policy.

Most Americans use AI but few will pay for it

Research from Tech Crunch found that only 2% of Americans were paying for AI services as of May, with an average cost of about $31 per month. The slow growth in paying users is pushing some AI companies to focus on business customers instead. AI companies face challenges because the technology is expensive to run. Both Republicans and Democrats are calling for more regulation of AI. President Trump has denied the need for AI safeguards, but Jay Clayton was named U.S. AI czar to oversee regulation policy.

Cato Networks Achieves AWS Security Competency for AI Security

Cato Networks earned the AWS Security Competency in AI Security for its Cato AI Security platform. The solution runs on AWS and helps secure AI built on AWS services like Amazon Bedrock and Amazon SageMaker. It inspects and protects traffic in real time without needing a separate security stack. Cato AI Security also helps prevent sensitive information from being shared with unapproved AI tools. Ofir Agasi, chief product officer at Cato Networks, said the competency recognizes the company's expertise in helping customers protect data and govern AI use at scale.

Cato Networks earns AWS AI Security Competency recognition

Cato Networks achieved an AWS Security Competency in AI Security. Its Cato AI Security tool runs on AWS and protects AI services like Amazon Bedrock and Amazon SageMaker in real time. It inspects traffic, governs activity, and helps prevent sensitive data from being shared with unapproved AI tools. Chief product officer Ofir Agasi said the recognition reflects Cato's expertise in helping customers secure AI on AWS.

Meta users frustrated by AI moderation mistakes and lack of human help

Some Facebook and Instagram users say Meta's AI moderation systems wrongly targeted their accounts and left them unable to get real help. Ronald Byrd lost monetization on his 1.3 million follower page. Melissa Mor, a motherhood jewelry business owner, had her account restored then shut down again. Both users hired attorneys and said they still had no resolution. Meta has not responded to multiple requests for comment.

Anthropic says AI agents did not breach Australian government websites

Anthropic's head of safeguards, Dave Orr, said investigations of hundreds of millions of transcripts found no unauthorized interactions with Australian government systems. He spoke during a joint parliamentary hearing on artificial intelligence. However, Orr acknowledged that Anthropic has limited visibility into customer usage because of standard zero data retention policies.

Survey finds AI misuse is top reputational threat for brands

A new Reputation Risk Index survey found that misuse of AI is the top reputational threat to brands. Former heads of state and senior executives from over 20 countries responded. They said AI shows promise for efficiency but also raises safety, data, fraud, and privacy concerns. The council advised business leaders to take immediate action to reduce AI-related brand damage.

FDA outlines 2027 plans for AI guidance in healthcare

The FDA is expected to release new AI guidance for healthcare in 2027. The guidance will cover AI use in clinical care, research, and healthcare administration. The FDA has worked with industry stakeholders and other agencies to align the guidance with healthcare needs. Utah is also expanding its AI sandbox for testing AI-powered healthcare tools.

Salesforce stock stays volatile as AI fears persist

Salesforce shares have been unstable in 2026 due to worries that AI agents could replace traditional software. The stock dropped over 40% by June but jumped 22% in August after strong Q2 earnings. It now sits near $230, still down about 10% from the start of the year. Options traders are using neutral strategies to profit from the ongoing uncertainty ahead of December earnings.

WorldTel launches AI and carbon technology services

WorldTel Corporation, based in Charlotte, North Carolina, launched enterprise technology services focused on artificial intelligence and carbon markets. The company operates across five divisions including AI, Sustainability, Digital Transformation, Technology and Platforms, and Human Capital. Co-founders Jason Roth and Richard Butler incorporated the company in North Carolina in 2025. WorldTel also formed the Ceario Carbon Exchange Alliance with InterstellarM2M in September 2026.

Explainable AI identifies best sales leads

Researchers at the University of Turku in Finland created a hybrid machine learning framework that predicts which B2B customers are likely to buy and explains why. The system combines SHAP explanations, self-organizing maps, and random forests to group customers and weigh contact decisions against business costs. The study used real sales records from a telecommunications company covering 3,007 customers and 53 features. The framework was published in the journal Applied Intelligence.

FEAgent checks if code is truly equivalent

FEAgent is a new tool that checks whether two programs are functionally equivalent by combining program-graph evidence with differential surrogate execution. It aligns public interfaces, issues bounded queries over different code relations, and uses blinded LLM surrogates to predict outcomes. A deterministic reconciler returns EQUIVALENT, INEQUIVALENT, or UNCLEAR when evidence conflicts. Testing on EquiBench and SWE-bench Verified showed that FEAgent found errors in benchmark labels and behavioral divergences missed by unit tests.

Sources

• Most Americans use AI, but few are willing to pay for it
• Most Americans use AI, but few are willing to pay for it
• Most Americans use AI, but few are willing to pay for it
• Most Americans use AI, but few are willing to pay for it
• Cato Networks Achieves AWS Security Competency in AI Security
• Cato Networks Achieves AWS Security Competency in AI Security
• Meta’s AI moderation system leaves users locked out, frustrated and searching for help
• Anthropic says AI agents didn’t breach Australian government websites
• Misuse of AI is brands’ top reputational threat, new survey says
• FDA spells out 2027 AI guidance plans
• Salesforce shares sit at the epicenter of AI confusion. Here's how to trade it
• WorldTel Corporation Launches Enterprise Technology Services Focused on Artificial Intelligence and Carbon Markets
• Explainable AI Maps the Hottest B2B Sales Leads Before Salespeople Dial
• Functionally Equivalent or Not? Graph-Grounded Differential Surrogate Execution for Code Equivalence