AI is reshaping both business opportunities and security risks in ways that demand attention. On the commercial side, companies like Oracle, Microsoft, and Anthropic are making bold moves to capture the growing demand for AI infrastructure and services. At the same time, cybersecurity threats are escalating fast, with AI-powered hackers and security breaches becoming more frequent and harder to detect.
Microsoft signed a 20-year energy deal with NiSource to secure roughly 1,100 megawatts of new generation capacity starting in 2028. The agreement also includes 600 MW of battery storage. NiSource expects the deal to boost savings for electric customers by about US$1.75 billion, or up to US$140 per customer annually. Microsoft is fully funding the generation and transmission infrastructure needed to support its data center operations.
Oracle is spending heavily on AI infrastructure, carrying a US$664 billion cloud backlog. The company plans to fund this through customer prepayments, supplier financing, and debt raises. However, rising credit default swap spreads and record borrowing costs create risk around Project Jupiter and related power projects. AI infrastructure may grow faster than actual workloads, potentially straining free cash flow before contracts ramp.
Oracle also launched a free AI tool called QuickStart to help early-stage startups manage finances. The tool connects startups to NetSuite and Oracle Cloud services from the start. Oracle hopes this will build long-term customer relationships while turning its large AI infrastructure contracts into profitable growth.
Rumble has shifted from a video platform into AI infrastructure, securing 250 MW of power from Northern Data and a six-year US$13.7 billion AI services contract reportedly with Anthropic. Analysts from Lake Street and Texas Capital both initiated coverage with Buy ratings, and Rumble's fair value estimate rose from US$22.00 to US$23.50.
Security concerns are mounting on multiple fronts. India's CERT-In warned that AI agents accessed Indian organizations' systems without permission, using exposed credentials from configuration files and public code repositories. The affected organizations did not detect the intrusions and only learned about them from AI model providers. CERT-In also found attackers using open source AI frameworks to speed up cyberattacks, reducing the time from access to damage from days to hours.
OpenCode, an open-source AI coding agent, disclosed a security breach affecting about 1.4% of its users. Attackers may have accessed API keys for OpenAI, Anthropic, and Google stored in the OpenCode Console dashboard. The breach originated from a legacy system being decommissioned. OpenCode is coordinating with Anthropic, OpenAI, and Google to fix the issue, and affected users have been asked to revoke and replace their API keys immediately.
Malwarebytes warns that AI assistants can be tricked by prompt injection attacks hidden in phishing emails. These attacks conceal harmful instructions that AI systems process while humans cannot read them. Experts say AI models struggle to tell normal text from hidden commands, and Malwarebytes recommends keeping humans in control, especially for high-impact actions like payments.
On the business adoption front, Benefit Street Partners uses OpenAI's Codex to analyze credit agreements and build internal tools faster. Portfolio manager Anant Kumar built a BDC mark tracker in about one day and a credit agreement analyzer that processes 300-page documents in roughly 20 minutes. The company uses GPT-6 Astra in a hosted portal but keeps human review for final decisions. Meanwhile, Don Joe describes ChatGPT as useful for answering questions and generating text, but notes it cannot think outside the box or replace human problem-solving.
AI-powered hackers attacks jumped 89% last year, pushing cybersecurity companies to rethink their defenses. Cybersecurity stocks are splitting into platforms and point solutions as the threat landscape evolves. Pagaya Technologies is expanding its AI-driven consumer lending platform, trading at $17.23 while some analysts estimate a fair value of $64.72. Q2 2026 results show the company has reached operating leverage and can generate cash even without a perfect economy.
Key Takeaways
- <ul><li>Microsoft secured 1,100 MW of power from NiSource for AI data centers through a 20-year deal starting in 2028, including 600 MW of battery storage.</li><li>Oracle carries a US$664 billion cloud backlog but faces risks from rising borrowing costs and potential cash flow strain as AI infrastructure outpaces workloads.</li><li>Oracle launched QuickStart, a free AI finance tool for startups, to build long-term customer relationships with small businesses.</li><li>Rumble secured a six-year US$13.7 billion AI services contract reportedly with Anthropic and 250 MW of power from Northern Data.</li><li>India's CERT-In warned that AI agents accessed organizational systems without permission using exposed credentials, and attackers used open source AI frameworks to speed up cyberattacks.</li><li>OpenCode disclosed a security breach exposing API keys for OpenAI, Anthropic, and Google for about 1.4% of its users.</li><li>Malwarebytes warns that AI assistants are vulnerable to prompt injection attacks hidden in phishing emails, and recommends keeping humans in control for high-impact actions.</li><li>Benefit Street Partners uses OpenAI's Codex to process 300-page credit agreements in about 20 minutes, but keeps human review for final decisions.</li><li>AI-powered hacker attacks jumped 89% last year, driving cybersecurity stocks to split into platforms and point solutions.</li><li>Pagaya Technologies trades at $17.23 while some analysts estimate fair value at $64.72, as its AI-driven lending platform reaches operating leverage.</li></ul>
3 Profitable AI Stocks Show Strong Earnings Growth
An article highlights three profitable AI companies including Lumentum Holdings, Ibiden, and Infomart. Lumentum makes optical chips for AI data centers. Ibiden produces circuit boards for GPUs and high performance servers. Infomart uses AI to turn faxes and PDFs into usable business data. These companies show how some AI businesses are already earning profits rather than just spending money.
Oracle Launches Free AI Finance Tool for Startups
Oracle released a free tool called QuickStart that uses AI to help early stage startups manage their finances. The tool connects startups to NetSuite and Oracle Cloud services from the start. Oracle hopes this will build long term customer relationships with small businesses. The move comes as Oracle works to turn its large AI infrastructure contracts into profitable growth.
Rumble Stock Fair Value Rises on AI Data Center Shift
Analysts raised their fair value estimate for RUM Group from US$22.00 to US$23.50. The change reflects Rumble's shift from a video platform into AI infrastructure. The company secured 250 MW of power from Northern Data and a six year US$13.7b AI services contract reportedly with Anthropic. Analysts from Lake Street and Texas Capital both initiated coverage with Buy ratings.
CERT-In Warns AI Agents Accessed Indian Systems Unauthorized
India's CERT-In issued a critical warning that AI agents accessed Indian organizations' systems without permission. The AI agents used exposed credentials from configuration files and public code repositories to access confidential business records. The affected organizations did not detect the intrusions and only learned about them from AI model providers. CERT-In also found attackers using open source AI frameworks to speed up cyberattacks, reducing the time from access to damage from days to hours.
OpenCode Reports Security Breach Exposing User API Keys
OpenCode, an open-source AI coding agent, disclosed a security breach affecting about 1.4% of its users. Attackers may have accessed API keys for OpenAI, Anthropic, and Google stored in the OpenCode Console dashboard. The breach originated from a legacy system being decommissioned. OpenCode is coordinating with Anthropic, OpenAI, and Google to fix the issue. Affected users have been asked to revoke and replace their API keys immediately.
Oracle faces AI data center delays despite massive cloud backlog
Oracle is spending heavily on AI infrastructure with a US$664 billion cloud backlog and strong Oracle Cloud Infrastructure growth. The company plans to fund this through customer prepayments, supplier financing, and debt raises. Rising credit default swap spreads and record borrowing costs create risk around Project Jupiter and related power projects. AI infrastructure may grow faster than actual workloads, which could strain free cash flow before contracts ramp.
AI agents fall for phishing attacks through hidden email tricks
Cybersecurity company Malwarebytes warns that AI assistants can be tricked by prompt injection attacks hidden in phishing emails. These attacks conceal harmful instructions that AI systems process while humans cannot read them. Experts say AI models struggle to tell normal text from hidden commands. Malwarebytes recommends keeping humans in control, especially for high-impact actions like payments, rather than letting AI agents act independently.
Microsoft secures 1,100 MW of power from NiSource for AI data centers
Microsoft signed a 20-year energy agreement with NiSource to supply roughly 1,100 megawatts of new generation capacity starting in 2028. Microsoft will fully fund the generation and transmission infrastructure needed for its data center demand. The deal also includes 600 MW of battery storage. NiSource expects the agreement to boost savings for electric customers to about US$1.75 billion, or up to US$140 per customer annually.
Benefit Street Partners uses Codex to speed up investment analysis
Benefit Street Partners uses OpenAI's Codex to analyze credit agreements and build internal tools faster. Portfolio manager Anant Kumar built a BDC mark tracker in about one day and a credit agreement analyzer that processes 300-page documents in roughly 20 minutes. The team now answers investment committee questions in a day instead of weeks. The company uses GPT-6 Astra in a hosted portal, but keeps human review for final decisions.
Don Joe explores ChatGPT usefulness and limits after Larry's introduction
Don Joe describes his experience using ChatGPT after being introduced to it by Larry. He finds the tool useful for answering questions and generating text. However, he notes that ChatGPT cannot think outside the box or come up with completely new ideas. He values it as help for his own problems but believes it does not replace human thinking or Larry's problem-solving skills.
How to Build a Smart AI Side Business
AI tools now make it easy to launch a business, but simply using AI does not guarantee success. One creator explains how Lovable turned an app idea into a profitable venture. The key is solving a real problem that people already understand. The article also highlights how Dropbox once spent too much to find customers before switching to a referral program. It stresses that knowing which work matters most can save months of effort.
Pagaya Technologies May Be Undervalued as AI Lending Grows
Pagaya Technologies is expanding its AI-driven consumer lending platform and adding new partners. The stock has fallen recently, trading at $17.23, while some analysts estimate a fair value of $64.72. Q2 2026 results show the company has reached operating leverage and can generate cash even without a perfect economy. Still, risks remain if fee margins shrink or funding for its loan network slows.
AI-Powered Hackers Are on the Rise as Cybersecurity Stocks Split
Cybersecurity stocks are splitting into platforms and point solutions as AI supercharges hackers. Attacks carried out by AI-powered hackers jumped 89% last year. This rapid growth is pushing companies to rethink how they defend against smarter and faster threats. The article explains how this shift is reshaping the cybersecurity market.
Sources
- 3 AI Stocks With Earnings Growth Of Up To 98%
- Oracle (ORCL) Rolls Out Free AI Finance Tool For Early Stage Startups
- Rumble (RUM) Stock Sees Modest Fair Value Lift As Analysts Back AI Shift
- AI Agents Accessed Indian Organizations Without Authorization, CERT-In Warns
- OpenCode Discloses Security Breach, Says Attackers May Have Accessed Users’ OpenAI, Anthropic And Google API Keys
- Learn Why The Bull Case For Oracle (ORCL) Could Change Following AI Data Center Delays
- Using an AI Agent Makes You a Sucker for Phishing Emails
- Microsoft Makes a Massive Move to Secure Its AI Future
- Benefit Street Partners answers questions faster with Codex
- SATURDAY MORNING WITH CRANKY DON JOE | Larry Introduced Me to Artificial Intelligence, and Now I Have Questions
- Starting an AI Side Hustle is Easy
- Is Pagaya Technologies (PGY) Undervalued On AI Lending Expansion And New Funding?
- As AI Supercharges Hackers, Cybersecurity Stocks Split Into Platforms and Point Solutions
Comments
Please log in to post a comment.