Amazon Spends $93M in US, Tech Capital Spending Hits $520B

Key Takeaways

• <ul><li>Amazon, Alphabet, Meta and Microsoft invested $269 billion of $520 billion in US capital spending last year</li><li>Amazon leads with over $93 million in domestic expenditures, up 61.1 percent from 2024</li><li>No federal law requires AI companies like Anthropic or OpenAI to disclose dangerous model behavior</li><li>DeepBrain AI launched emotionally responsive AI Roleplay training on September 15, 2026</li><li>Connecticut regulates AI in health insurance for 270,000 enrollees, effective January 1, 2027</li><li>Tech sector productivity grew 61.3 percent while manufacturing and utilities fell</li><li>Oracle made the top 10 with more than $15.7 million in domestic investments</li><li>Connecticut's rules ban exclusive AI use for reducing payments and protect member data from training models</li><li>Senate lawmakers are considering new AI disclosure rules</li><li>Bank of Georgia automated 95% of unsecured loans, approved in under 15 seconds</li></ul>

AI STUDIOS Adds Emotional AI Roleplay for Sales Training

AI STUDIOS, DeepBrain AI's enterprise platform, announced enhanced emotional expression capabilities for AI Roleplay on September 15, 2026. The interactive training solution lets employees practice realistic customer conversations with AI customers that show emotions through facial expressions and tone of voice. Companies can customize training with their own product information and customer profiles. The solution is already used in insurance sales, contact centers, in-store customer service and wealth management training.

DeepBrain AI Expands AI Roleplay Across Sales and Wealth Management

DeepBrain AI announced expanded emotional expression capabilities for AI Roleplay through its AI STUDIOS platform on September 15, 2026. The company plans to expand the training solution across sales, customer service, wealth management and other customer-facing teams. AI customers now express emotions and react differently based on personalities and attitudes, creating more realistic training experiences. The solution provides real-time feedback and analysis of speaking pace, intonation and facial expressions.

AI Roleplay Training Expands With Emotionally Responsive AI Customers

AI STUDIOS, DeepBrain AI's enterprise AI video and avatar platform, launched enhanced emotional expression features for AI Roleplay on September 15, 2026. Employees can practice full customer conversations with AI customers that respond based on personality, interests and attitudes rather than predetermined scripts. The solution analyzes participant performance and provides coaching on speaking pace, intonation and facial expressions. It is already used in insurance sales training, contact center training and financial advisory team rehearsals.

AI Buildout Pushes Tech Firms to Top of US Investment Ranking

A Progressive Policy Institute report found that Amazon, Alphabet, Meta and Microsoft invested $269 billion of the $520 billion in US capital spending last year. Amazon leads with over $93 million in estimated domestic expenditures, a 61.1 percent increase from 2024. The tech firms' spending surge was driven by increased demand for compute and AI infrastructure. Oracle also made the top 10 with more than $15.7 million in domestic investments.

Tech Giants Dominate US Capital Spending in AI Study

The Progressive Policy Institute's annual Investment Heroes report, set for release on September 16, 2026, shows Big Tech firms dominate US investment rankings. Amazon, Alphabet, Meta and Microsoft accounted for more than half of all US capital spending last year at $269 billion. The tech sector's productivity grew 61.3 percent, while manufacturing and utilities sectors fell. The report credits the AI boom and data center buildout for driving this growth.

AI companies face no federal law requiring disclosure of dangerous incidents

There is no U.S. federal law requiring AI companies to disclose dangerous incidents involving their models. Some companies like DeepMind have voluntarily disclosed issues, while others like Meta have faced criticism for not doing so. Industry groups are working toward transparency standards. The lack of clear rules has created a patchwork of disclosure practices across the industry.

No single U.S. law requires AI companies to report dangerous model behavior

No single federal law targets AI companies like Anthropic or OpenAI regarding disclosure of dangerous model behavior. Existing laws cover some areas like cybersecurity breaches and data breaches in healthcare or finance. The FTC can act on deceptive practices and the Justice Department can use fraud statutes. A company discovering alarming AI behavior in testing may have no obligation to disclose it without a breach or harm trigger. Senate lawmakers are considering new rules.

Arteris launches FlexGen Multi-Die to advance AI chiplet architecture

Arteris expanded its multi-die portfolio with FlexGen Multi-Die, a product designed to help semiconductor companies shift from monolithic chip designs to chiplet-based architectures for AI and high-performance computing. The product tackles efficient and secure data movement across multiple chips. It gives engineering teams more automation and architectural flexibility. Arteris believes the product is well-positioned to meet growing demand in the semiconductor industry.

Bank of Georgia uses AI-native infrastructure to transform digital banking

Bank of Georgia won nine awards at the 2026 World's Best Digital Banks by building customer-centric AI-native infrastructure. 95% of its unsecured loans are automated and approved in under 15 seconds. Its AI assistant resolves almost 80% of customer queries at 91% satisfaction. The bank's app functions as a financial super-app integrating taxes, utilities, and vehicle ownership services. It also cut bank guarantee processing from several days to under an hour.

Connecticut announces AI regulations for state employee health plans

Connecticut Comptroller Sean Scanlon announced regulations governing AI use by health insurance companies managing the state employee health plan. The rules apply to over 270,000 enrollees and include a ban on exclusive AI use for reducing payments and a guarantee that member data will not train AI models. The regulations take effect January 1, 2027. Scanlon plans to recommend lawmakers extend these protections to all state-regulated plans in 2027. Providers including Anthem, Cigna, Aetna, and Caremark serve the state plan.

UiPath PATH Stock Falls as Traders Doubt AI Growth

UiPath stock fell more than 3 percent on September 16, 2026, amid concerns about its AI growth story. BofA raised its price target from $13 to $15 but kept an Underperform rating. CEO Daniel Dines sold about 1.4 million shares for roughly $22.5 million. The company reported quarterly revenue near $410 million with strong margins but faces questions about future AI-driven growth.

Build Serverless PII Redaction Pipeline Using Amazon Bedrock

Amazon Bedrock Data Automation can help organizations redact personally identifiable information from thousands of scanned documents automatically. The solution uses a custom blueprint and a serverless pipeline with AWS Step Functions and AWS Lambda. It handles fields like patient names, dates of birth, and addresses while preserving non-sensitive data. This approach reduces human error and improves compliance compared to manual redaction.

Slower AI Race May Still Lead to Next Profit Wave

Some AI leaders, including Anthropic CEO Dario Amodei and OpenAI CEO Sam Altman, have called for slowing frontier AI development. Analyst Luke Lango says investors may still profit because businesses are only beginning to use existing AI. He notes inference could make up about two-thirds of AI computing in 2026. He also highlights a food-service robotics startup using current AI in real-world settings.

Priority Pass Gets £500 Million Investment for Digital and AI Growth

Priority Pass received a £500 million investment from Collinson Group to expand its digital and AI capabilities. The funding will also grow its lounge network and travel services. Priority Pass already gives members access to over 1,300 airport lounges worldwide. The investment aims to improve the traveler experience and strengthen its position in the loyalty and rewards industry.

Sources

• AI STUDIOS Expands Sales Training With Emotionally Expressive AI Roleplay
• AI STUDIOS Expands Sales Training With Emotionally Expressive AI Roleplay
• AI STUDIOS Expands Sales Training With Emotionally Expressive AI Roleplay
• AI buildout propels tech firms to top of US investment ranking: Study
• AI Buildout Propels Tech Firms to Top of US Investment Ranking: Study
• Explainer-Do AI companies have to disclose dangerous incidents?
• Do AI companies have to disclose dangerous incidents?
• Is Arteris (AIP) Using FlexGen Multi-Die to Quietly Reshape Its AI Chiplet Investment Story?
• Bank of Georiga: AI-Native Infrastructure And The Future Of Digital Banking
• CT announces AI regulations for state employee health plans
• UiPath PATH Stock Slumps As Traders Question AI Growth Story
• Build a serverless PII redaction pipeline with Amazon Bedrock Data Automation
• If the Race for Smarter AI Slows, This May Be the Next Wave of Profits
• Priority Pass steps up digital, AI and lounge expansion under £500 million Collinson investment plan