AI Industry News — September 26, 2026

Findings

1. [1] 2026-09-26 Meta says its rejection of ads promoting the new documentary about Elon Musk "was an error and the ads are being restored"; YouTube is also allowing the ads (Anthony Ha/TechCrunch)

   Anthony Ha / TechCrunch: Meta says its rejection of ads promoting the new documentary about Elon Musk “was an error and the ads are being restored”; YouTube is also allowing the ads  —  At least two companies appear to be reversing their stance on whether to accept advertising for director Alex Gibney's upcoming documentary about Elon Musk.

2. [2] 2026-09-26 A US federal jury awarded Taction Technology $5.7B+ in damages, finding Apple infringed haptics patents; Apple to appeal, says it "does not use Taction's" tech (Ashley Capoot/CNBC)

   Ashley Capoot / CNBC: A US federal jury awarded Taction Technology $5.7B+ in damages, finding Apple infringed haptics patents; Apple to appeal, says it “does not use Taction's” tech  —  A federal jury in San Diego awarded Taction Technology more than $5.7 billion in damages Friday after finding …

3. [3] 2026-09-26 Bitget CEO Gracy Chen says she suspects North Korean attackers exploited a backend system used to process wallet transactions to drain $387M from the platform (Camila Grigera Naón/Fortune)

   Camila Grigera Naón / Fortune: Bitget CEO Gracy Chen says she suspects North Korean attackers exploited a backend system used to process wallet transactions to drain $387M from the platform  —  Another major hack has rattled the cryptocurrency world.  On Thursday night, crypto exchange Bitget suffered a security breach …

4. [4] 2026-09-26 Two-thirds of IT leaders report AI results, but few would interrupt the CEO's vacation over them

   Speaking to 160 IT vice presidents in Las Vegas, tech entrepreneur Azeem Azhar asked who had measurable AI results. Two-thirds raised their hands. Then he asked who had results good enough to interrupt the CEO's summer vacation. Only eight did. Whether that slow pace of progress justifies the massive investments is one of the key questions in the AI bubble…

5. [5] 2026-09-26 Docker Cloud Sandboxes Provide a Consistent Sandbox Abstraction Across Laptop and Cloud

   Docker Cloud Sandboxes provide secure, hosted execution environments for running AI coding agents on Docker-managed infrastructure. Built on hardware-enforced microVM isolation, the platform provides a consistent execution environment and unified CLI workflows for seamlessly moving workloads from local machines to the cloud. By Sergio De Simone

6. [6] 2026-09-26 AI access makes people almost entirely unwilling to say "I don't know," study finds

   A study with more than 3,000 participants shows that just having access to AI answers nearly eliminated people's willingness to say "I don't know." In one experiment, it dropped from 44 to 3 percent, even though the AI was almost always wrong. Participants who used AI felt more confident but were correct only about a third as often as those…

7. [7] 2026-09-26 Brazil's Lula da Silva signs an EO banning online betting; app stores and telecom providers must block platforms on Oct. 6; Congress has 120 days to approve it (Reuters)

   Reuters: Brazil's Lula da Silva signs an EO banning online betting; app stores and telecom providers must block platforms on Oct. 6; Congress has 120 days to approve it  —  Brazil's President Luiz Inacio Lula da Silva on Friday announced a ban on online betting operations in Latin America's largest economy …

8. [8] 2026-09-26 Sources: YouTube, TikTok, and Meta's Instagram and Facebook declined to run paid ads for Alex Gibney's Musk documentary, citing "political content" (Steven Zeitchik/The Hollywood Reporter)

   Steven Zeitchik / The Hollywood Reporter: Sources: YouTube, TikTok, and Meta's Instagram and Facebook declined to run paid ads for Alex Gibney's Musk documentary, citing “political content”  —  YouTube, TikTok and Meta have all turned down the trailer for the Oct. 9 doc, citing “political” content.  X won't even take Bleecker Street's calls.

9. [9] 2026-09-26 Sources: Automattic CEO Matt Mullenweg announces a new board and new advisers; board members include two founders of IRL and the author of the Silo book series (Sarah Perez/TechCrunch)

   Sarah Perez / TechCrunch: Sources: Automattic CEO Matt Mullenweg announces a new board and new advisers; board members include two founders of IRL and the author of the Silo book series  —  Automattic CEO Matt Mullenweg has rebuilt the company's board just weeks after its previous directors tried, and failed, to oust him.

10. [10] 2026-09-26 Sources: Monzo is in talks to be acquired by Nubank parent Nu Holdings in a deal valuing it at £8B-£10B; Monzo also discussed selling a ~15% stake to PE firms (Laith Al-Khalaf/Financial Times)

    Laith Al-Khalaf / Financial Times: Sources: Monzo is in talks to be acquired by Nubank parent Nu Holdings in a deal valuing it at £8B-£10B; Monzo also discussed selling a ~15% stake to PE firms  —  UK fintech is also in discussions with private equity about selling a significant stake in the business

11. [11] 2026-09-26 The US and China create a "Super Intelligence Dialogue" on AI risks and a separate AI incident hotline, likened to a Cold War-era "red telephone" (Ben Berkowitz/Axios)

    Ben Berkowitz / Axios: The US and China create a “Super Intelligence Dialogue” on AI risks and a separate AI incident hotline, likened to a Cold War-era “red telephone”  —  The U.S. and China have agreed to set up a dialogue on artificial intelligence, with a communications channel …

12. [12] 2026-09-26 Walmart CEO John Furner says the company won't use its AI shopping assistant or electronic shelf labels to change product prices based on a shopper's identity (Gregory Meyer/Financial Times)

    Gregory Meyer / Financial Times: Walmart CEO John Furner says the company won't use its AI shopping assistant or electronic shelf labels to change product prices based on a shopper's identity  —  Largest US retailer issues open letter saying it will not use electronic shelf labels to change costs based on shopper identity

13. [13] 2026-09-26 Presentation: Adaptive Recommenders in the Real World: Inference, Evals, and System Design

    Mallika Rao explains that the true complexity of adaptive recommendation systems lies outside model architecture. She discusses how real-time feedback loops, retrieval freshness, multi-stage orchestration, and end-to-end latency budgeting enable systems to continuously learn and evolve in production under real-world operational constraints like latency, cost, and observability. By Mallika Rao

14. [14] 2026-09-26 Nvidia's SoL-Pi system cuts coding agent token usage nearly in half by optimizing the harness

    SoL-Pi cuts coding agents' token usage by up to 49 percent with little change in performance by optimizing the control layer between the model and its environment. A research agent tested 152 approaches across more than 3,000 runs to develop the system, though the gains were smaller on other benchmarks. The article Nvidia's SoL-Pi system cuts coding agent token usage…

15. [15] 2026-09-26 PitchBook: VCs have invested $4B+ in quantum computing companies YTD, almost as much as in all of 2025, which nearly matched the previous four years combined (Financial Times)

    Financial Times: PitchBook: VCs have invested $4B+ in quantum computing companies YTD, almost as much as in all of 2025, which nearly matched the previous four years combined  —  The dawn of a new computing era may finally be here.  Now the race is on to find a path to profit.

16. [16] 2026-09-26 OpenAI's GPT-6 Astra can now tell you exactly where you screwed up your IKEA shelf

    OpenAI's GPT-6 Astra can look at a photo and tell whether an IKEA furniture piece was assembled incorrectly, hitting an 80 percent accuracy rate. Back in November 2025, the best model managed just 28 percent. According to Epoch AI, the speed isn't quite fast enough yet for real-time assembly guidance, but the gap is closing quickly. The article OpenAI's GPT-6…

17. [17] 2026-09-26 OpenAI pauses its "most capable models" after agents exploit loopholes and leak data

    OpenAI has shared new details from its ongoing AI safety investigation. One research model exploited a DNS loophole to reach the internet from a locked-down environment, while another deliberately leaked a GitHub token and twice ignored a researcher's direct instructions. OpenAI has paused tool-based training, evaluation, and inference for its most capable models. With government and university sites among those…

18. [18] 2026-09-26 A look at the wave of Google DeepMind researchers who have exited recently to launch their own AI startups focused on alternatives to LLMs (Bloomberg)

    Bloomberg: A look at the wave of Google DeepMind researchers who have exited recently to launch their own AI startups focused on alternatives to LLMs  —  When a group of 15 Google DeepMind employees and alumni met for a breakfast this month in central London, the conversation quickly turned …

19. [19] 2026-09-26 NYC-based Confido, a provider of AI-powered workflow automation tools for consumer packaged goods companies, raised a $55M Series B led by Insight Partners (AlleyWatch)

    AlleyWatch: NYC-based Confido, a provider of AI-powered workflow automation tools for consumer packaged goods companies, raised a $55M Series B led by Insight Partners  —  Consumer packaged goods brands operate on razor-thin margins, yet the finance, accounting, trade spend, and operations work that protects …

20. [20] 2026-09-26 Russia has increased targeted strikes on Ukrainian data centers, disrupting internet access for ~100K Kyiv residents on Wednesday and Thursday (Christopher Miller/Financial Times)

    Christopher Miller / Financial Times: Russia has increased targeted strikes on Ukrainian data centers, disrupting internet access for ~100K Kyiv residents on Wednesday and Thursday  —  Kyiv residents experience two days of internet disruptions, raising fears over access to banking and other online services

21. [21] 2026-09-26 Palantir and 8VC cofounder Joe Lonsdale, an investor in Anthropic, says AI companies are attempting to sway public policy by warning of existential AI risks (Joe Brock/Reuters)

    Joe Brock / Reuters: Palantir and 8VC cofounder Joe Lonsdale, an investor in Anthropic, says AI companies are attempting to sway public policy by warning of existential AI risks  —  Joe Lonsdale, an investor in Anthropic and the co-founder of technology company Palantir (PLTR.O), said on Friday that efforts …

22. [22] 2026-09-26 A US federal judge dealt fresh setbacks to Deel in the Rippling case over alleged spying, including rejecting its bid to strike testimony from a central witness (Peter Blumberg/Bloomberg)

    Peter Blumberg / Bloomberg: A US federal judge dealt fresh setbacks to Deel in the Rippling case over alleged spying, including rejecting its bid to strike testimony from a central witness  —  Deel Inc. was hit with new setbacks in Rippling Inc.'s corporate espionage litigation against the HR startup.

23. [23] 2026-09-26 What to expect at Dell’s AI Leadership Symposium: Join theCUBE Sept. 29

    Enterprise AI is moving past experimentation, and the infrastructure decisions behind it are becoming business decisions. As companies put AI into production, the conversation is expanding beyond models and GPUs to the data, infrastructure and operating models required to support AI at scale. Dell Technologies Inc. has been building around that shift through its AI […] The post What to…

24. [24] 2026-09-25 What to expect at NetApp INSIGHT: Join theCUBE Sept. 30

    As artificial intelligence goes into production, unified storage could be the glue that keeps everything together. NetApp Inc., a leader in enterprise data storage, wants to lift the burden of AI on enterprise infrastructure. Data management has proven to be one of the biggest hurdles for AI adopters, so NetApp’s unified storage solutions aims to […] The post What to…

25. [25] 2026-09-25 4 insights from Dreamforce: AI agents move from demos to measurable outcomes

    At Dreamforce 2026, AI agents are being judged less on what they can say and more on what they get done. Across Salesforce Inc.’s ecosystem, the standalone toolkits that let technology teams build quick proofs of concept are giving way to integrated platforms. These connect agents to enterprise data, to workflows and to the people […] The post 4 insights…

26. [26] 2026-09-25 Microsoft overhauls Copilot with new coding, document editing features

    Microsoft Corp. today launched a new version of Copilot, the artificial intelligence assistant included in its Microsoft 365 productivity suite. The biggest addition to the tool is an interface section called Code. It enables nontechnical workers to create simple apps using prompts. For example, a business analyst could create a dashboard that tracks store sales. […] The post Microsoft overhauls…

27. [27] 2026-09-25 Vanderbilt University extends identity governance to AI agents

    Identity governance must accommodate changing relationships among people, institutions and technology. Vanderbilt University sees that complexity deepening as artificial intelligence agents enter institutional workflows. Vanderbilt uses OneVU, powered by Okta, for single sign-on. Its broader environment includes residential life, athletics, its own police department and more than $1 billion in research. That breadth complicates access […] The post Vanderbilt University…

Sources

1. [1] Meta says its rejection of ads promoting the new documentary about Elon Musk "was an error and the ads are being restored"; YouTube is also allowing the ads (Anthony Ha/TechCrunch)
2. [2] A US federal jury awarded Taction Technology $5.7B+ in damages, finding Apple infringed haptics patents; Apple to appeal, says it "does not use Taction's" tech (Ashley Capoot/CNBC)
3. [3] Bitget CEO Gracy Chen says she suspects North Korean attackers exploited a backend system used to process wallet transactions to drain $387M from the platform (Camila Grigera Naón/Fortune)
4. [4] Two-thirds of IT leaders report AI results, but few would interrupt the CEO's vacation over them
5. [5] Docker Cloud Sandboxes Provide a Consistent Sandbox Abstraction Across Laptop and Cloud
6. [6] AI access makes people almost entirely unwilling to say "I don't know," study finds
7. [7] Brazil's Lula da Silva signs an EO banning online betting; app stores and telecom providers must block platforms on Oct. 6; Congress has 120 days to approve it (Reuters)
8. [8] Sources: YouTube, TikTok, and Meta's Instagram and Facebook declined to run paid ads for Alex Gibney's Musk documentary, citing "political content" (Steven Zeitchik/The Hollywood Reporter)
9. [9] Sources: Automattic CEO Matt Mullenweg announces a new board and new advisers; board members include two founders of IRL and the author of the Silo book series (Sarah Perez/TechCrunch)
10. [10] Sources: Monzo is in talks to be acquired by Nubank parent Nu Holdings in a deal valuing it at £8B-£10B; Monzo also discussed selling a ~15% stake to PE firms (Laith Al-Khalaf/Financial Times)
11. [11] The US and China create a "Super Intelligence Dialogue" on AI risks and a separate AI incident hotline, likened to a Cold War-era "red telephone" (Ben Berkowitz/Axios)
12. [12] Walmart CEO John Furner says the company won't use its AI shopping assistant or electronic shelf labels to change product prices based on a shopper's identity (Gregory Meyer/Financial Times)
13. [13] Presentation: Adaptive Recommenders in the Real World: Inference, Evals, and System Design
14. [14] Nvidia's SoL-Pi system cuts coding agent token usage nearly in half by optimizing the harness
15. [15] PitchBook: VCs have invested $4B+ in quantum computing companies YTD, almost as much as in all of 2025, which nearly matched the previous four years combined (Financial Times)
16. [16] OpenAI's GPT-6 Astra can now tell you exactly where you screwed up your IKEA shelf
17. [17] OpenAI pauses its "most capable models" after agents exploit loopholes and leak data
18. [18] A look at the wave of Google DeepMind researchers who have exited recently to launch their own AI startups focused on alternatives to LLMs (Bloomberg)
19. [19] NYC-based Confido, a provider of AI-powered workflow automation tools for consumer packaged goods companies, raised a $55M Series B led by Insight Partners (AlleyWatch)
20. [20] Russia has increased targeted strikes on Ukrainian data centers, disrupting internet access for ~100K Kyiv residents on Wednesday and Thursday (Christopher Miller/Financial Times)
21. [21] Palantir and 8VC cofounder Joe Lonsdale, an investor in Anthropic, says AI companies are attempting to sway public policy by warning of existential AI risks (Joe Brock/Reuters)
22. [22] A US federal judge dealt fresh setbacks to Deel in the Rippling case over alleged spying, including rejecting its bid to strike testimony from a central witness (Peter Blumberg/Bloomberg)
23. [23] What to expect at Dell’s AI Leadership Symposium: Join theCUBE Sept. 29
24. [24] What to expect at NetApp INSIGHT: Join theCUBE Sept. 30
25. [25] 4 insights from Dreamforce: AI agents move from demos to measurable outcomes
26. [26] Microsoft overhauls Copilot with new coding, document editing features
27. [27] Vanderbilt University extends identity governance to AI agents