google, microsoft and deepseek Updates

Recent developments in artificial intelligence highlight both rapid innovation and emerging challenges across various sectors. On November 26, 2025, Waton Financial Limited launched its new AI investment app, TradingWTF. This platform utilizes the company's DePearl multi-agent architecture and autonomous AI Traders to offer fast, informed trading, including a one-click Copy Trade feature. Chairman Kai Zhou stated that TradingWTF aims to democratize institutional-grade market insights, supported by a partnership with Panda AI, and Waton Financial plans to introduce subscription services for market analysis. Also on November 26, Google introduced its new AI image generation and editing tool, Nano Banana Pro, which immediately raised concerns among experts due to the highly realistic nature of the images it produces. However, the rapid pace of AI deployment also brings security vulnerabilities. Israeli cybersecurity researchers from Cato Networks' Cato CTRL group uncovered a significant flaw, dubbed "HashJack," affecting AI browsers like Google's Gemini, Microsoft's Copilot, and Perplexity's Comet. This technique hides malicious instructions in a website's URL, potentially leading AI assistants to display fake information or steal user data. While Microsoft and Perplexity quickly issued fixes for Copilot for Edge and Comet, Google has not yet resolved the issue for Gemini in Chrome, considering it "intended behavior" as of November 25, 2025. Separately, Google's new AI coding tool, Antigravity, was hacked within 24 hours of its launch, with security researcher Aaron Portnoy identifying a flaw that could allow malware installation on Windows and Mac computers. The global competition in AI, particularly between the US and China, is intensifying, with a significant focus on power resources. China's DeepSeek AI model demonstrates efficiency with less computing power, while the US prioritizes building AI Data Centers. Experts project these data centers could consume 20% of global electricity by 2030-2035, driven by demands from training large language models, AI inference, multimodal AI like OpenAI's Sora, and autonomous vehicles, potentially leading to higher consumer power costs. This energy demand underscores the infrastructure challenges accompanying AI's expansion. Economically, AI continues to reshape industries. HSBC predicts the S&P 500 index will reach 7,500 by the end of 2026, a 40% increase over two years, largely driven by the AI boom's ability to boost productivity and corporate earnings. Conversely, McKinsey is cutting approximately 200 tech jobs as it shifts more tasks to artificial intelligence, with further reductions anticipated over the next two years. CEO Julie Sweet noted the company is letting go of staff who cannot be retrained for AI-related work, reflecting a broader trend of job displacement in sectors like banking. Government and enterprise adoption of AI is also accelerating. Enabled Intelligence, a startup founded in 2020, secured a substantial $708 million contract from the National Geospatial-Intelligence Agency (NGA) for its Sequoia program. This initiative will use data labeling to train AI computer vision models to identify targets in satellite and airborne images for defense and intelligence purposes. In the healthcare and life science sectors, Merck KGaA launched a new high-performance computing platform, a supercomputer built with Lenovo and housed in an Equinix data center in Munich, to accelerate AI and data research, including drug discovery and material modeling, while also meeting environmental goals. Meanwhile, Florida Insurance Commissioner Michael Yaworsky is advocating for regulators to oversee AI use in insurance, emphasizing responsible governance, human oversight, and the disclosure of AI applications, with lawmakers proposing bills to ensure human final decisions on claim denials.

Key Takeaways

• Waton Financial launched its AI investment app, TradingWTF, on November 26, 2025, utilizing DePearl multi-agent architecture and a partnership with Panda AI.
• A "HashJack" security flaw was discovered in AI browsers like Google's Gemini, Microsoft's Copilot, and Perplexity's Comet, allowing malicious URL instructions.
• While Microsoft and Perplexity fixed the HashJack flaw, Google has not resolved it for Gemini in Chrome as of November 25, 2025, and its new AI coding tool, Antigravity, was hacked within 24 hours of launch.
• Google's Nano Banana Pro, an AI image generation tool launched November 26, 2025, raises concerns due to its highly realistic image output.
• The global AI race, including China's DeepSeek model, demands significant power, with experts predicting data centers could consume 20% of global electricity by 2030-2035.
• McKinsey is cutting approximately 200 tech jobs and plans further reductions as it shifts tasks to AI, reflecting a broader industry trend of AI-driven job displacement.
• HSBC predicts the S&P 500 will reach 7,500 by the end of 2026, a 40% increase, driven by the AI boom's impact on productivity and corporate earnings.
• Enabled Intelligence won a $708 million contract from the NGA for its Sequoia program, using AI data labeling to train computer vision models for defense and intelligence.
• Merck KGaA launched a new high-performance computing platform, a supercomputer built with Lenovo, to accelerate AI and data research in life science, healthcare, and electronics.
• Florida regulators are seeking oversight for AI use in insurance, emphasizing responsible governance, human oversight, and disclosure, with lawmakers proposing related bills.

Waton Financial launches AI trading app TradingWTF

Waton Financial (NASDAQ: WTF) launched its new AI-powered investment app, TradingWTF, on November 26, 2025. This app uses the company's DePearl multi-agent architecture to offer autonomous AI Traders. It includes a one-click Copy Trade feature, allowing users to easily copy algorithmic strategies. TradingWTF aims to make institutional-level trading strategies available to more investors and operates 24/7 in real time. The company also has a strategic partnership with Panda AI for this initiative.

Waton Financial introduces AI trading platform TradingWTF

Waton Financial Limited (NASDAQ: WTF) launched its new AI investment app, TradingWTF, on November 26, 2025. This platform uses the company's DePearl multi-agent architecture and autonomous AI Traders to offer fast and informed trading. It features a one-click Copy Trade function, letting users easily follow AI-driven strategies. Chairman Kai Zhou stated that TradingWTF aims to make institutional-grade insights available to everyone and is supported by a partnership with Panda AI. Waton Financial plans to add subscription services for market analysis and reports in the future.

Waton Financial unveils AI investment app TradingWTF

Waton Financial Limited (NASDAQ: WTF) launched its new AI-powered investment app, TradingWTF, on November 26, 2025. The app uses the company's DePearl multi-agent architecture and autonomous AI Traders to provide fast and smart trading experiences. It includes a Copy Trade feature that allows users to easily replicate advanced algorithmic strategies. Chairman Kai Zhou said this launch helps democratize market insights and is part of their goal to become an AI-agents holding company, supported by a partnership with Panda AI. Waton plans to introduce subscription services for stock analysis and market monitoring.

Israeli firm finds AI browser security flaw

Israeli cybersecurity researchers from Cato Networks' Cato CTRL group found a major security flaw in AI browsers like Google's Gemini, Microsoft's Copilot, and Perplexity's Comet. This new "HashJack" technique uses malicious instructions hidden in a website's URL after the hashtag symbol. Attackers can use this to make AI assistants show fake information, steal user data, or trick users into bad decisions. While Microsoft and Perplexity have fixed the issue in Copilot for Edge and Comet, Google has not yet resolved it for Gemini in Chrome as of November 25, 2025.

New HashJack attack hijacks AI browsers

Security researchers at Cato Networks discovered a new attack called HashJack that can take over AI browsers and assistants. This technique hides harmful instructions in the hashtag part of a website's address. It can trick tools like Perplexity's Comet, Microsoft's Copilot for Edge, and Google's Gemini for Chrome into showing fake links or stealing private information. While Microsoft and Perplexity have released fixes, Google considers it "intended behavior" or social engineering, not a security flaw. The attack requires users to interact with the AI assistant, making widespread exploitation less likely.

The AI race demands huge power resources

The global competition in AI, especially between the US and China, is becoming a race for power. China's DeepSeek AI model has shown it can achieve results with less computing power, challenging US leadership. The US is focusing on building more AI Data Centers, which are essential but consume vast amounts of electricity. Experts predict data centers could use 20% of global electricity by 2030-2035, leading to significant increases in power costs for consumers. This high energy demand comes from training large language models, AI inference, multimodal AI like OpenAI's Sora, AI-driven research, and autonomous vehicles.

Google's Nano Banana Pro raises image realism concerns

Google launched its new AI image generation and editing tool called Nano Banana Pro on November 26, 2025. Experts are worried because the images created by Nano Banana Pro look too real. NBC News reporter Erin McLaughlin tested the new AI generator. A content creator is also teaching people how to spot these fake images.

McKinsey cuts 200 tech jobs moves to AI

McKinsey is cutting about 200 tech jobs as it moves more tasks to artificial intelligence. The company is considering more job reductions across different areas over the next two years as it increases AI use. A spokesperson stated that AI offers great opportunities for McKinsey and its clients, making support functions more efficient. Global Managing Partner Bob Sternfels focuses on client-facing roles, while CEO Julie Sweet noted they are letting go of staff who cannot be retrained for AI-related work. This trend reflects broader industry changes, with banks also expecting significant job cuts due to AI.

HSBC predicts S&P 500 will reach 7500 by 2026

HSBC predicts the S&P 500 index will climb to 7,500 by the end of 2026, a 40% increase over two years. Bill Dudley, HSBC's U.S. equity strategist, believes the ongoing artificial intelligence boom will drive this growth. He noted that AI's rapid adoption and its ability to boost productivity will lead to strong corporate earnings. While acknowledging a "K-shaped economy" where some consumers struggle, Dudley is confident the AI revolution will overcome these challenges for the overall market.

Enabled Intelligence wins NGA's 708 million AI contract

Enabled Intelligence, a startup founded in 2020, won a $708 million contract from the National Geospatial-Intelligence Agency (NGA). This contract is for the NGA's Sequoia program, which will use data labeling to train AI computer vision models. These models will help find targets of interest in satellite and airborne images for the Department of War and the Intelligence Community. Enabled Intelligence specializes in AI data labeling for classified systems and partners with companies like BAE Systems and Vantor.

Florida seeks rules for AI use in insurance

Florida Insurance Commissioner Michael Yaworsky wants regulators to oversee how insurance companies use artificial intelligence, stressing the need for responsible AI governance. Lawmakers are also addressing the issue, with Representative Cassel and Senator Bradley filing bills to ensure humans make final decisions on insurance claim denials. Yaworsky emphasized the importance of disclosing AI use, conducting audits, and having human oversight. House Speaker Paul Renner announced an "Artificial Intelligence Week" to discuss AI's impact across various sectors. Yaworsky highlighted a case where a health company used an AI solution without understanding its workings, showing the need for clear regulations.

Merck KGaA launches supercomputer for AI research

Merck KGaA launched a new high-performance computing platform to boost its AI and data research. This supercomputer, built with Lenovo and housed in an Equinix data center in Munich, will improve research and development in life science, healthcare, and electronics. Laura Matz, Chief Science and Technology Officer, stated it will accelerate drug discovery and material modeling. The platform uses Lenovo's Neptune liquid cooling for efficiency and combines private and public cloud resources. Equinix provides the digital foundation, allowing Merck KGaA to run complex AI models and simulations while meeting its environmental goals.

Google's new AI coding tool hacked quickly

Google's new AI coding tool, Antigravity, was hacked within 24 hours of its launch. Security researcher Aaron Portnoy found a serious flaw that could allow attackers to install malware on users' Windows and Mac computers. This "backdoor" could spy on victims or run ransomware if a user clicked "trusted" on malicious code. Portnoy reported the issue to Google, but no fix was available at the time. Experts like Gadi Evron warn that many AI coding tools are released quickly without enough security testing, making them vulnerable targets for hackers.

Sources

• Waton Financial (NASDAQ: WTF) unveils TradingWTF DePearl AI copy trading app
• Waton Financial Limited Launches TradingWTF, a DePearl™ Powered AI Trading Platform Designed to Reshape How Investors Trade
• Waton Financial Limited Launches AI-Powered Investment App TradingWTF to Democratize Trading Access
• AI Browser Security Flaw Discovered by Israeli Researchers
• New "HashJack" attack can hijack AI browsers and assistants
• The AI Race Is Becoming A Sprint For Power
• Google's Nano Banana Pro is raising concerns over realistic AI image generation
• McKinsey Cuts About 200 Tech Jobs, Shifts More Roles to AI
• HSBC sees S&P 500 hitting 7,500 by end of 2026 with 'more to come' in the AI trade
• Startup Enabled Intelligence nabs NGA's $708 million AI training contract
• Florida wants regulators to oversee insurance companies' use of artificial intelligence: "Responsible AI governance is crucial"
• Merck KGaA: The Supercomputer for AI and Data Research
• Google’s hot new AI coding tool was hacked a day after launch