KnowBe4 Launches AI Tool Control Browser Extension

Key Takeaways

• KnowBe4 launched a browser extension giving security teams control over unapproved AI tool usage, as 52% of organizations report ungoverned AI use.
• Deloitte estimates UK employees will spend one billion pounds on unregulated AI tools, posing serious IP and data security risks.
• Concentrix reported a $988 million net loss on $2.45 billion revenue due to a goodwill impairment, despite over half of revenue coming from AI services.
• Sharplink rebranded from SharpLink Gaming to focus on Ethereum treasury management and AI agent digital asset management.
• New research introduces the Observation-Aliasing Impossibility Theorem explaining why AI agents fail when collaborating as teams.
• Nvidia approved a $150 billion share repurchase increase while Micron beat expectations and OpenAI launched dots. to compete with Meta's Muse.
• Salesforce rose 46% in Q3 with AI tools like Claudeforce, while Palantir and Microsoft also gained strongly.
• Meta launched Muse AI agent for car shopping, with Carvana as a top pick due to its owned logistics network and record Q2 sales of 197,000 units.
• California subpoenaed OpenAI over AI model security incidents, including a model hacking Hugging Face.
• Kevin Mandia's Armadin raised $255.5 million at a $2.5 billion valuation, using agentic swarms for security testing.

KnowBe4 launches browser extension to control unapproved AI use

KnowBe4 added a browser extension for Chrome and Edge to its Agent Risk Manager product. It gives security teams real-time control over employees using AI tools without approval. The extension can identify access to known AI sites and let administrators permit, warn or block that access. Research cited by KnowBe4 found 52% of organizations said their AI use was unapproved or ungoverned. The extension is available to certain customers through an early access program.

UK workers expected to spend one billion pounds on unregulated AI tools

A Deloitte study found that UK employees are expected to spend an estimated one billion pounds of their own money on unregulated, privately funded AI tools. The report calls shadow AI a serious operational and strategic security risk for CIOs and CISOs. The main risk is the uncontrolled loss of intellectual property, source code, financial data or personal information. The article advises organizations to offer secure alternatives, set dynamic governance policies, improve technical detection and invest in training.

Concentrix reports AI growth and large goodwill impairment in Q3

Concentrix reported third quarter revenue of US$2.45 billion and a net loss of US$988.11 million due to a large non cash goodwill impairment. More than half of Concentrix's quarterly revenue now comes from AI related services. Management guided to lower constant currency revenue for the rest of the fiscal year and an operating loss for 2026. Non GAAP operating margins and earnings remained positive despite pressure from reduced work with two large hyperscale clients.

Sharplink pivots to Ethereum treasury and AI wallet vision

Sharplink rebranded from SharpLink Gaming and refocused on institutional-grade Ethereum treasury management. The CEO also described a vision for AI agents managing digital assets through dedicated blockchain wallets and smart contracts. The article says the company's investment narrative now depends on whether its Ethereum treasury platform can become a trusted institutional business. Fair value estimates from the Simply Wall St community range from about US$14.38 to US$42.49.

Study examines why AI agents can fail when working as a team

The paper studies the global coherence problem, where AI agents each make locally valid decisions but jointly produce an invalid result. It introduces the Observation-Aliasing Impossibility Theorem to explain the exact boundary of this failure. On a controlled revision benchmark, a frontier model scored 40 out of 40 when the deciding event was visible but only 12 to 17 out of 40 when it was hidden. The research concludes that local intelligence cannot replace missing global state.

Micron beats expectations while Nvidia announces massive buyback

Micron Technology reported strong Q4 results that eased worries about the AI infrastructure boom slowing down. Nvidia's board approved a $150 billion increase to its share repurchase program, the largest in the company's history. OpenAI launched dots. to compete with Meta's Muse AI assistant. Anthropic reported $42 billion in losses on $4.6 billion in revenue. Trump and AI CEOs agreed to a voluntary AI safety framework.

Jim Cramer says software stock bets paid off in Q3

Jim Cramer said investors who bet that AI would destroy software stocks lost money. Salesforce rose about 46% in Q3 with new AI tools like Claudeforce. Veeva Systems soared about 60% and Palantir also performed well. Microsoft gained about 37% during the quarter. Cramer said software bounced back while some data center stocks faced profit-taking.

AI car shopping agents could help Carvana gain market share

Meta Platforms launched Muse, an AI agent that can browse websites and negotiate prices for users. Morgan Stanley analysts say car buying is a prime use case for these agents. Carvana is the top pick because it owns its reconditioning and logistics network. Carvana's Q2 retail sales hit a record of about 197,000 units. Morgan Stanley rates Carvana Overweight with a $90 price target.

California subpoenas OpenAI over AI model security incidents

California Attorney General Rob Bonta issued a subpoena to OpenAI. The investigation follows incidents involving the company's AI models, including one where a model hacked AI startup Hugging Face. Bonta said companies have a legal responsibility to ensure their models do not enable cyberattacks. OpenAI said it looks forward to providing information and has strengthened safeguards since the incident.

Kevin Mandia's security startup Armadin raises $255.5 million

Kevin Mandia's new startup Armadin raised $255.5 million in a Series B round at a $2.5 billion valuation. Andreessen Horowitz and Accel led the round. Armadin uses agentic swarms to continuously test security by chaining together vulnerabilities to hack into systems. The company raised over $445 million total, including a $190 million Series A in March.

AI Creates Verification Bottlenecks in Companies

AI is creating bottlenecks in companies by making verification the most important step. Managers still need to check AI outputs and catch errors that automated systems miss. Recording these correction moments could train AI models to reduce the need for human oversight. In the future, this automation might even remove the need for company hierarchy altogether.

Nova Sees Long-Term Growth in AI Chip Packaging

Nova provides measurement equipment for semiconductor manufacturing and maintains a competitive advantage. As of October 1, 2026, the stock trades at $386 with a Superscore of 88 out of 100. The company's platforms serve as the tool of record at leading foundries for advanced packaging and Gate-All-Around nodes. Revenue growth is driven by the shift toward more measurement-intensive manufacturing nodes, though high valuation multiples and customer concentration pose risks.

Sources

• KnowBe4 adds browser extension to curb shadow AI use
• The billion-pound vacuum: Why shadow AI is forcing businesses to act
• Concentrix Stock Sees AI Growth Offset Goodwill Hit In Q3
• Should Sharplink’s Ethereum Pivot and AI Wallet Vision Reshape the Core SBET Investment Narrative?
• Global Coherence: When Every Agent Is Right and the Team Is Still Wrong - A Local-to-Global Semantic Foundation for Multi-Agent Collaboration
• Stocktwits AI Roundup: Micron’s Blowout Quarter, Nvidia’s $150B Buyback And The Race For AI Agents
• Jim Cramer Says Investors Who Bet AI Would Destroy Software Stocks ‘Blew Up’
• AI could haggle for your next car, and one stock stands to win
• California attorney general subpoenas OpenAI over incidents involving its AI models
• Kevin Mandia's new 'agent swarm' security startup Armadin raises $255.5M at $2.5B valuation
• AI Is Making Verification the Bottleneck for Companies
• Nova's 2026 Outlook: Tool-of-Record Status Drives Long-Term Growth in AI Chip Packaging