Chicago Public Schools’ limits on student use of ChatGPT, Claude and Google Gemini are colliding with Google’s AI-enabled Search. Google says schools can opt out of Gemini and other AI tools, but it cannot disable AI Mode and AI Overviews inside Search; parents and students protested outside its Fulton Market office. CPS is trying to suppress those features and may look for another search engine, although the district also depends on Google Classroom, Drive, Docs and Chromebooks. The district cites privacy and safety concerns, has issued guidance emphasizing human oversight, and offers examples such as using AI to make study guides. Some parents and teachers remain concerned about learning and screen time.
Enterprise AI spending is also shaping the fortunes of hardware suppliers. Dell has more than 6,500 AI customers, including 3,300 added over three quarters. In fiscal 2027’s second quarter, it reported $46.97 billion in revenue, up 57.8% year over year, record AI-server orders of $60.9 billion and raised its full-year revenue outlook to $192 billion. But memory and processor shortages, higher capacity investment and a 47.22% drop in free cash flow to $986 million are pressures. A 24/7 Wall St. analysis assigned Dell a Buy rating and a $434.72 target, compared with a listed price of $335.75.
Seagate reported fiscal first-quarter revenue of $4.233 billion, up 61%, with data centers contributing $3.4 billion, or 81% of sales. Microsoft qualified a solution using Seagate’s HAMR Mozaic platform. Seagate shares nevertheless fell 20% in the week after results. The company expects HAMR to account for about half of nearline exabyte output by the end of 2026 and plans to set prices for calendar-2028 supply over the next few quarters—a test of pricing power amid customer concentration, component-cost and data-center-spending risks.
Anthropic is changing how its Claude Code teams manage increasingly autonomous work. Engineer Lydia Hallie says the company ships more than 30,000 changes a day; Opus 55 is her daily model and Fable 52 checks its output. Auto mode, the default since August, uses a classifier to approve routine work after deny rules are checked. Hallie recommends operating-system sandboxes, clear task tests and consistent CI checks. In hooks, exit code 2 sends Claude back to continue work, while code 1 marks a task complete; Code projects remains in beta as the orchestrator model.
AI trading tools and cyber operations raise separate oversight concerns. Forex brokers can struggle to distinguish trades placed by agents connected through the open MCP standard, including Claude, because orders appear ordinary. The article warns that agent errors or unreliable third-party information can trigger real trades, and that poorly secured MCP servers may expose account credentials or tokens. It recommends more frequent client-profile reviews, monitoring changes in trading and asking platform providers to label MCP orders. Separately, Palo Alto Networks’ Unit 42 says Chinese-speaking attackers used a customized Cairn AI penetration-testing framework in 25 operations against at least 19 organizations in August and September 2026, gaining access or taking data from 10. The framework used Zhipu GLM 5.2 and DeepSeek V4 Flash and V4 Pro; targets included semiconductor, manufacturing, defense and human-rights groups. Unit 42 assessed the activity with moderate confidence as China-linked intelligence gathering and cited familiar weaknesses such as poor passwords, authentication flaws, exposed cloud APIs and unpatched software.
Nvidia’s leadership uncertainty and broader AI-investment concerns have rattled markets. One report says Nvidia shares plunged 13%, a semiconductor gauge fell 3.5% and the Nasdaq 100 declined 1.4%; it also reports a 5% drop after CEO Jensen Huang was ousted. Nvidia’s market value fell below $4 trillion. The board is seeking a permanent replacement, with finance chief Colette Kress serving as interim CEO; AMD and Broadcom gained as the S&P 500 edged higher. A separate NBC News report, published Nov. 12, says analysts see Nvidia’s fundamentals as intact but notes the company has never operated without Huang. Investor Ken Mahoney warned of near-term volatility.
Investors are also trimming exposure to AI shares: LPL Research says hedge-fund exposure fell about 8% from its June peak, with managers selling holdings about 3.5 times faster than they opened new short positions. For the week ending Nov. 6, information technology fell 8.2% and the Russell 1000 Growth index lost 4.8%. LPL recommends watching whether exposure continues to shrink and market gains broaden beyond a few AI leaders, while long-term investors should consider concentration in popular stocks. Meanwhile, Fivetran research finds data work is consuming AI resources: 39% of companies spend at least a quarter of AI budgets on data-quality fixes, integration rebuilding and reprocessing, and 58% of technical teams spend at least a quarter of their AI time on such avoidable work. Though 83% invested at least $1 million in AI last year, only 37% have multiple projects running at scale. For the next year, more respondents plan to prioritize data infrastructure (31%) than talent (17%) or AI models (16%).
A graduate-seminar pilot offers a different measure of AI’s role. Students studying weaknesses in European integration first developed their own ideas, then used AI for a presentation and explained how they used the tools. Researchers found AI accelerated routine tasks and let students explore more ambitious methods, while the assignment kept independent analysis and judgment central.
Key Takeaways
- Google says schools can opt out of Gemini and other AI tools, but cannot turn off AI Mode and AI Overviews within Search; Chicago Public Schools is trying to suppress them and may consider another search engine.
- CPS blocks student access to ChatGPT, Claude and Gemini over privacy and safety concerns, while its guidance emphasizes human oversight; the district also relies on Google’s education services and Chromebooks.
- Dell reported $46.97 billion in fiscal 2027 second-quarter revenue, record AI-server orders of $60.9 billion and a raised $192 billion full-year outlook, while facing component shortages and lower free cash flow.
- Seagate’s revenue rose 61% to $4.233 billion, with data centers accounting for 81%; Microsoft qualified a solution based on Seagate’s HAMR Mozaic platform, and 2028 supply pricing is an upcoming test.
- Anthropic says it ships more than 30,000 changes daily and uses Claude Code controls including Auto mode, deny rules, sandboxes and CI checks to manage autonomous coding work.
- AI agents connected to trading accounts through MCP can be difficult for brokers to identify; the article flags trade errors and credential exposure risks and recommends monitoring and labeling MCP orders.
- Unit 42 attributed 25 Cairn framework operations against at least 19 organizations to Chinese-speaking attackers, with access or data theft at 10; the customized framework used DeepSeek V4 models and Zhipu GLM 5.2.
- Nvidia’s leadership uncertainty coincided with reported share declines and a market value below $4 trillion; Colette Kress is interim CEO as the board seeks a permanent replacement, while AMD and Broadcom gained in one report.
- LPL Research says hedge-fund AI-stock exposure is down about 8% from June’s peak; Fivetran reports that data remediation absorbs at least a quarter of AI budgets at 39% of surveyed companies.
- A graduate seminar pilot found AI sped routine work and supported more ambitious analysis when students first formed their own ideas and explained their AI use.
Chicago families protest Google’s AI search policy for schools
Chicago parents and students protested outside Google’s Fulton Market office on Wednesday after the company said schools that block its AI search features could lose access to Google Search. Google plans to turn off Gemini and other AI tools for schools unless administrators opt in, but says it cannot disable AI Mode and AI Overview within Search. Chicago Public Schools has blocked ChatGPT, Claude and the Gemini app on student accounts and devices, and is trying to suppress AI features in Search. The district may consider another search engine if it cannot keep Google Search without AI. Google says its Guided Learning feature will offer step by-step help rather than direct answers, but some parents and teachers remain concerned about AI’s effect on student learning.
Why Chicago students could lose Google Search over AI restrictions
Chicago Public Schools blocks student access to ChatGPT, Claude, Google Gemini and other generative AI tools because of privacy and safety concerns. The district has also issued guidance for teachers and students on using approved AI tools, with human oversight and examples such as making study guides. Google’s AI search features could put students’ access to Google Search at risk if CPS restricts them. The district is exploring ways to suppress those features and says it may seek another search service if it cannot provide Google Search without AI. CPS also relies on Google Classroom, Drive, Docs and Chromebooks, so finding alternatives would be a major undertaking.
Chicago schools weigh AI rules as Google Search access faces risk
Chicago Public Schools does not allow students to use popular generative AI tools such as ChatGPT, Claude and Google Gemini because of privacy and safety concerns. The district has issued guidance for educators and students over the past two years on using AI in teaching and learning, with a focus on human oversight. Its guidance offers examples such as creating study guides and generating images related to books. Parents have raised concerns about AI’s effects on learning and want to limit students’ screen time.
Dell’s expanding enterprise AI business could lift its stock
Dell Technologies shares have fallen sharply, but 24/7 Wall St. set a $434.72 price target, about 29.5% above the listed current price of $335.75, and gave the stock a Buy rating. Dell’s AI customer base has grown to more than 6,500, with 3,300 added in the last three quarters, as enterprise and government buyers expand their AI spending. In its second quarter of fiscal 2027, Dell reported $46.97 billion in revenue, up 57.8% year over year, and adjusted earnings per share of $7.04. AI server orders reached a record $60.9 billion, and the company raised its full-year revenue outlook to $192 billion. Dell also faces shortages of memory and processors, while free cash flow fell 47.22% to $986 million as capital spending rose to build AI capacity.
Seagate’s AI demand story faces a test in 2028 pricing talks
Seagate Technology reported a 61% rise in fiscal first-quarter revenue to $4.233 billion and beat earnings estimates for the sixth straight quarter. Its data center sales reached $3.4 billion, or 81% of revenue, and Microsoft qualified a solution based on Seagate’s HAMR Mozaic platform. Despite the results, Seagate shares fell 20% in the week after the earnings report and traded at about 13 times forward earnings. Management plans to set prices for supply allocated for calendar 2028 over the next few quarters, a key test of the company’s pricing power. Seagate expects HAMR to make up about half of its nearline exabyte output by the end of calendar 2026, while risks include customer concentration, rising component costs and possible delays in data center spending.
Anthropic updates Claude Code for more autonomous work
Anthropic engineer Lydia Hallie said the company now ships more than 30,000 changes each day, driving updates to Claude Code. Opus 55 is her daily model, while Fable 52 checks its own output. Auto mode became the default in August, and a classifier approves routine work after Claude Code checks deny rules. Hallie said teams should use operating system sandboxes to block access to sensitive files and unapproved sites, and define tasks with clear tests. Hooks must exit with code 2 to send Claude back for more work, while exit code 1 still marks a task complete. CI runs the same checks on every pull request, and Code projects remains in beta as the orchestrator model.
AI trading agents leave brokers unable to identify who is placing trades
Forex brokers can usually identify automated trades made with Expert Advisors, but AI agents connected through the open MCP standard are harder to spot. Traders can connect agents such as Claude to live accounts using official or open-source MCP servers, whose safety controls vary. AI mistakes or false information from third-party tools can lead to real trades, while brokers see only ordinary orders. Hosted or unchecked servers may also expose account logins or access tokens to theft or misuse. The article says brokers should review client profiles more often, watch for changes in trading and related orders, and ask platform providers to label MCP orders.
AI hacking framework helped attackers breach 10 organizations
Chinese-speaking attackers used the open-source Cairn AI penetration testing framework in 25 operations against at least 19 organizations in August and September 2026. Palo Alto Networks’ Unit 42 said the attackers gained access or took data from 10 organizations, and assessed with moderate confidence that the activity was China-linked intelligence gathering. A customized version of Cairn ran from a Hong Kong server and used Zhipu GLM 5.2, DeepSeek V4 Flash, and DeepSeek V4 Pro. Targets included semiconductor and manufacturing groups, defense organizations, and human rights groups; compromised information included more than one million teachers’ union member records and about 500 user records at a defense-policy program. Unit 42 said the attackers exploited familiar weaknesses such as poor passwords, authentication flaws, exposed cloud APIs, and unpatched software, while AI increased the number of targets they could work on.
Nvidia leadership turmoil shakes AI stocks as shares plunge
Nvidia shares fell sharply amid a leadership crisis, while investors reconsidered spending on AI infrastructure. The article reports that the company’s shares plunged 13%, a semiconductor gauge dropped 3.5%, and the Nasdaq 100 fell 1.4%; it also says Nvidia fell 5% after CEO Jensen Huang was ousted. Nvidia’s market value dropped below $4 trillion. The S&P 500 edged higher, with Advanced Micro Devices and Broadcom among the gainers. While the board seeks a permanent replacement for Huang, longtime finance chief Colette Kress is serving as interim CEO, and investor Ken Mahoney warned of near-term volatility.
Nvidia’s future raises questions for the AI economy
An NBC News article published November 12, 2026, examines what comes next for Nvidia, a company at the center of the AI economy. Analysts said the company’s fundamentals remain intact and the broader market has held up. However, Nvidia has never operated without its founder, Jensen Huang, who was shown as CEO in Tokyo in July. The article’s available text does not provide further details about the company’s plans or leadership.
Graduate students test how AI can support independent analysis
A pilot study in a graduate seminar examined whether AI can expand students’ analytical skills without replacing their judgment. The seminar focused on weaknesses in European integration. An “AI-enriched contract” asked students to develop their own ideas before using AI, give an AI-assisted presentation, and explain how they used the tools. The study found that AI sped up routine tasks and allowed students to explore more ambitious methods.
Hedge funds cut AI stock holdings as investors weigh market risks
Hedge funds have reduced their exposure to AI stocks, but the selling appears to reflect risk reduction rather than a prediction that the AI boom will fail. LPL Research said hedge fund exposure had fallen about 8% from its June peak, while managers sold stocks they owned about 3.5 times faster than they opened new short positions. For the week ending Nov. 6, the information technology sector fell 8.2% and the Russell 1000 Growth index dropped 4.8%. LPL advises watching whether hedge fund exposure keeps shrinking and whether gains spread beyond a few AI leaders. It also says long-term investors should consider how much of their portfolios depend on a small group of popular stocks.
Fivetran finds data problems consume a large share of AI budgets
Fivetran research found that 39% of companies spend at least a quarter of their AI budgets fixing data quality, rebuilding integrations, and processing data again. Although 83% of surveyed organizations invested at least $1 million in AI last year, only 37% have multiple AI projects running at scale. Technical teams also spend time on avoidable work, with 58% devoting at least a quarter of their AI time to issues such as remediation and rebuilding integrations. Companies that rely heavily on custom integrations are more likely to lose at least a quarter of their AI budget and face data-related delays longer than a month than those using standardized platforms. For the next 12 months, 31% plan to prioritize data infrastructure, ahead of talent at 17% and AI models at 16%.
Sources
- Chicago parents, students protest Google’s new no-AI, no-search policy with schools
- What to know about CPS and AI as Chicago students risk losing access to Google search
- What to know about CPS and AI as Chicago students risk losing access to Google search
- Prediction: One AI Trend Could Send Dell Stock Higher
- Seagate’s Legitimacy as a Play on AI Hardware Demand is About To Be Tested
- Anthropic retools Claude Code for hands-off runs
- Same Login, Same Orders, Different Trader: Brokers Can't Tell When AI Trades Live
- Attackers breached 10 organizations with an open-source AI pentest framework
- Nasdaq 100 falls 1.4% as Nvidia crisis rattles AI trade
- Nvidia is at the center of the AI economy. What happens now?
- AI – Master and Commander: Lessons from graduate student projects and the need for an AI-enriched contract
- Hedge Funds Are Dumping AI Stocks—Here’s Why Retail Investors Shouldn’t Panic
- Fivetran: AI budget consumed by fixing data issues
Comments
Please log in to post a comment.