AI-linked shares came under pressure ahead of Nvidia’s earnings, dragging U.S. markets lower. Nvidia fell 4.5% in premarket trading, while futures pointed to declines across the Nasdaq, S&P 500 and Dow. Investors also weighed concerns about rising interest rates and crowded AI trades. A separate report said the Nasdaq lost $2.5 trillion in market value on November 5, 2026; Nvidia shares fell 20% during the month and shed $573 billion in value over the preceding week. Reports that CEO Jensen Huang had resigned added to investor concern, while Alphabet’s DeepMind and Microsoft’s Nuance Communications also declined.
Companies are also working to make AI deployment more verifiable and economical. Detectify introduced Agentic Tooling so AI agents can test live applications for exploitable vulnerabilities using its testing engines, attack-surface data and payload assessments. Its append-only Security Ledger records requests, payloads and results, and the tools run in customers’ environments through the Model Context Protocol or a command-line interface. Potentially harmful tests are flagged for human review. Available on request from November 16, 2026, the product addresses a concern Detectify says customers have raised: about 80% of findings from AI-driven static application security testing tools are noise.
Anthropic updated its Claude models for coding, reasoning and cybersecurity. The company says Mythos 5.2 detected 10.2% more vulnerabilities than Claude Opus 5.5 and 14.4% more than Mythos 5.1 in its CVE Rediscovery evaluation; a security company reported that one in six confirmed bugs it found was missed by every other frontier model tested. Mythos 5.2 is available to selected partners, while Fable 5.2 is for regular users. Anthropic says Fable’s safeguards block high-risk activity such as penetration testing and exploit generation, but it reduced unnecessary cybersecurity blocks by about 60% versus Fable 5.1. The company also tested Mythos against sandbox escapes and cautioned that evaluations may not catch every failure.
Microsoft plans to unveil Copilot and Azure features at its Ignite conference, starting Tuesday, including an Azure system designed to lower customers’ AI-running costs. The Information said the plans are intended to help Microsoft compete with Anthropic and other AI companies. Meanwhile, Gartner reports that the cost of achieving measurable AI returns has risen 18-fold in two years. In a second-quarter 2026 survey of more than 1,000 functional leaders, 52% reported some positive results from AI projects. Nearly three-fourths of more than 2,000 CIOs and technology executives planned to deploy agentic AI within 12 months, but only 14% of CIOs in India and Asia-Pacific had agentic-AI governance capabilities in place. Gartner urged organizations to tie projects to business results and build oversight.
AI’s effects are also showing up in company valuations. A European Internet Ventures report says 10 listed companies across travel search, price comparison, creative software, freelance marketplaces and online education lost $206 billion in combined market value from December 2021 to November 2026, falling from $305 billion to $99 billion. Tripadvisor lost 77%, while Shutterstock and Chegg each lost 98%; Shutterstock’s third-quarter 2026 revenue was down 33% year over year. Coursera’s valuation fell 73%, and Stack Overflow’s new questions dropped 99% between November 2022 and September 2026. The report argues that businesses completing transactions or delivering services may be more resilient than platforms focused mainly on information or leads.
In model infrastructure, MLCommons’ MLPerf Training v6.1 results, announced November 16, 2026, introduce recommendation-system and large-scale AI post-training benchmarks. The submissions included 153 systems and 250 results, with 12 systems using at least 1,000 accelerators and one using 10,240. Tests ran on hardware from Nvidia, AMD and Google, including Nvidia Rubin GPUs and Google TPUv7. Separately, SoundHound AI reported record third-quarter 2026 revenue of $62.4 million, up 48.5% year over year and above the $61.71 million expected by Wall Street. Its adjusted loss was $0.05 per share, versus an expected $0.13 loss. Analysts’ average $7.10 price target implies 103.4% upside, even as the stock is down 65% year to date.
Key Takeaways
- Nvidia fell 4.5% in premarket trading ahead of earnings; a separate November 5 selloff erased $2.5 trillion from Nasdaq market value, while Nvidia lost $573 billion over the preceding week.
- AI-stock declines weighed on U.S. markets amid interest-rate concerns and warnings about crowded AI trades; Alphabet’s DeepMind and Microsoft’s Nuance Communications also fell.
- Detectify’s Agentic Tooling lets AI agents test live applications and log evidence in an append-only Security Ledger; potentially harmful tests require human review. Availability on request begins November 16, 2026.
- Anthropic says Mythos 5.2 found 10.2% more vulnerabilities than Claude Opus 5.5 and 14.4% more than Mythos 5.1 in its evaluation. Fable 5.2 reduces unnecessary cybersecurity blocks by about 60% versus Fable 5.1.
- Microsoft plans Copilot and Azure announcements at Ignite, including an Azure system designed to lower AI operating costs as it competes with Anthropic and other AI companies.
- Gartner says the cost of achieving measurable AI returns rose 18-fold in two years; 52% of surveyed functional leaders reported some positive results, while only 14% of CIOs in India and Asia-Pacific had agentic-AI governance capabilities.
- A European Internet Ventures report estimates 10 companies in five AI-exposed sectors lost $206 billion in market value from December 2021 to November 2026; Shutterstock and Chegg each fell 98%.
- MLPerf Training v6.1 recorded 153 systems and 250 results, including a 10,240-accelerator system, using Nvidia, AMD and Google hardware such as Nvidia Rubin GPUs and Google TPUv7.
- SoundHound AI posted Q3 2026 revenue of $62.4 million, up 48.5% year over year, and an adjusted loss of $0.05 per share; its stock is down 65% year to date.
Nvidia slide pressures AI stocks ahead of earnings
U.S. stock futures pointed lower on Monday as investors sold shares in artificial intelligence and semiconductor companies ahead of Nvidia’s earnings later in the week. Nvidia fell 4.5% in premarket trading, while futures indicated a 0.4% drop for the Nasdaq and a 0.2% drop for the S&P 500. The article linked Nvidia’s decline to concerns after CEO Jensen Huang announced his resignation last week.
AI stock losses pull Wall Street lower at week’s start
U.S. stock futures fell as AI shares declined ahead of a busy week for technology companies. In premarket trading, the Dow Jones Industrial Average fell 0.2%, the S&P 500 dropped 0.3%, and the Nasdaq Composite fell 0.4%. The article said investors were watching the AI sector because it has driven market growth in recent years and losses could affect the broader market.
Nvidia and other AI stocks drag U.S. market lower
AI stocks fell on Monday, pulling the U.S. market lower, with Nvidia leading the decline. Nvidia dropped 4.5% in premarket trading, while shares of Alphabet’s DeepMind and Microsoft’s Nuance Communications also fell. The article said concerns about rising interest rates were weighing on the technology sector, and the Dow Jones Industrial Average and S&P 500 were expected to open lower. It also noted that Nvidia’s recent weakness could affect other AI stocks and add to market volatility.
Detectify launches tools to verify AI security investigations
Detectify launched Agentic Tooling to help DevSecOps teams check whether software vulnerabilities can be exploited. AI agents can use Detectify’s testing engines, built over more than a decade, to test live applications from the outside and confirm findings. The company says customers report that about 80% of findings from AI-driven static application security testing tools are noise. Detectify’s append-only Security Ledger records requests, payloads, results, and links to the original ticket so teams can review the evidence behind an agent’s conclusion. The tools are available on request starting November 16, 2026.
Detectify gives AI agents tools to test live application flaws
Detectify introduced Agentic Tooling for DevSecOps teams that want evidence behind security findings from AI agents. Agents can use the company’s testing engines, attack surface data, and payload assessments to test live applications from the outside. The Security Ledger records each test’s requests, payloads, and results in an append-only log written by Detectify, helping teams separate verified findings from unsupported AI conclusions. The tools run in a customer’s environment and connect through the Model Context Protocol or a command-line interface. Tickets for tests that could cause harm, such as denial-of-service tests, are marked for a person to decide on, and the product is available on request starting November 16, 2026.
SoundHound AI beats Q3 estimates as analysts see 103% upside
SoundHound AI reported record revenue of $62.4 million in Q3 2026, up 48.5% from a year earlier and above the $61.71 million expected by Wall Street. Its adjusted loss was $0.05 per share, smaller than the expected loss of $0.13. Analysts give the U.S. voice AI company a Moderate Buy rating and an average price target of $7.10, which suggests 103.4% upside from current levels. SOUN shares are down 65% year to date.
Nasdaq loses $2.5 trillion as NVIDIA fears shake AI stocks
On November 5, 2026, the Nasdaq saw $2.5 trillion in market value disappear during a sharp market session. The selloff followed Lakewood Capital Management’s Anthony Bozza warning that investors were crowding into the same AI trades. NVIDIA shares fell, and reports that CEO Jensen Huang had resigned added to investor concerns. NVIDIA’s stock dropped 20% during the month, losing $573 billion in market value over the past week. The turmoil made investors more cautious about AI stocks and led them to seek safer ways to handle technology market swings.
AI returns cost 18 times more as businesses expand adoption
Gartner says the cost of achieving measurable returns from AI has risen 18-fold in two years, even as companies increase their use of the technology. A second-quarter 2026 survey of more than 1,000 functional leaders found that 52% had seen some positive results from AI projects. Another Gartner survey of more than 2,000 CIOs and technology executives found that nearly three-fourths planned to deploy agentic AI within 12 months. Yet only 14% of CIOs in India and the Asia-Pacific region had governance capabilities for agentic AI in place. Gartner urged companies to connect AI projects to business results and build oversight as they adopt the technology, since the financial gains may also go to customers, vendors, or outsourcing partners.
Anthropic reduces Claude Fable 5.2 security blocks as Mythos finds more bugs
Anthropic rolled out Claude Mythos 5.2 to selected partners and Fable 5.2 to regular users, with updates to coding and reasoning. Mythos 5.2 detected 10.2% more vulnerabilities than Claude Opus 5.5 and 14.4% more than Mythos 5.1 in Anthropic’s CVE Rediscovery evaluation. A security company also reported that one in six confirmed bugs found by Mythos 5.2 had been missed by every other frontier model it tested. Fable 5.2 includes safeguards against high-risk activities such as penetration testing and exploit generation, but Anthropic says it reduced unnecessary cybersecurity blocks by about 60% compared with Fable 5.1. Anthropic also tested Mythos 5.2 against sandbox escape attempts and said its evaluations may not identify every possible failure before release.
AI disruption cuts $206 billion from companies in five sectors
A European Internet Ventures report published on November 16 says 10 listed companies in AI-exposed sectors lost $206 billion in market value between December 2021 and November 2026. Their combined value fell from $305 billion to $99 billion, across travel search, price comparison, creative software, freelance marketplaces, and online education. Tripadvisor lost 77% of its market value, while Shutterstock and Chegg each lost 98%; Shutterstock’s third-quarter 2026 revenue fell 33% year over year. The report links these pressures to AI tools that can compare travel options, create content, or help students with homework. Coursera’s valuation fell 73%, and Stack Overflow recorded a 99% drop in new questions between November 2022 and September 2026. The report says businesses that complete transactions or deliver services may be more resilient than platforms that mainly provide information or leads.
MLPerf Training v6.1 adds new AI benchmarks and record participation
MLCommons announced MLPerf Training v6.1 results on Nov. 16, 2026, with new benchmarks for recommendation systems and large-scale AI post-training. DLRMv4-HSTU replaces the earlier DLRM-dcnv2 benchmark and trains on the Yambda dataset. The Qwen3.5-397B benchmark tests post-training with Group Relative Policy Optimization to help systems learn tasks such as tool-calling. Participants submitted 153 systems and 250 results, including 12 systems with at least 1,000 accelerators and one with 10,240. The submissions used NVIDIA, AMD, and Google hardware, including NVIDIA Rubin GPUs and Google TPUv7.
Microsoft plans Azure features to lower AI costs
Microsoft plans to introduce new features for Copilot and Azure at its Ignite conference, which starts Tuesday. The Information reported that the features aim to help Microsoft compete with Anthropic and other AI companies. One planned Azure system is designed to reduce customers’ costs of running AI.
Sources
- S&P 500, Nasdaq set for lower open as AI stocks feel Nvidia heat
- AI Stocks Weigh on Wall Street to Start the Week
- More drops for Nvidia and other AI stocks drag the U.S. market lower
- Detectify launches Agentic Tooling to bring verifiable evidence to AI-led security investigations
- Detectify's Agentic Tooling lets AI agents test suspected flaws in running applications
- SoundHound AI Stock Forecast: How Much Upside Do Analysts See Post-Q3 Earnings?
- Nasdaq session erases $2.5 trillion amid NVIDIA AI trading panic
- AI returns become harder to achieve as business investment costs surge
- Anthropic’s Claude Fable 5.2 blocks fewer security tasks, while Mythos gets better at finding bugs
- Which businesses are being disrupted by AI? From travel platforms to education, here’s what is changing
- MLCommons Releases MLPerf Training v6.1 Results, Spotlighting Recent Advances in AI Training
- Microsoft to Debut Products Aimed at Reducing Customers’ AI Costs
Comments
Please log in to post a comment.