Anna E. Molosky

Something was different at #NYTechWeek this year—in the best way possible. Over the week of dinners, demonstrations, and conversations, leaders moved beyond pilots, focusing instead on how AI is creating measurable ROI and how to govern agents at scale. A few standouts: 👇 ———— Thanks for organizing @a16z, @andrewchen, @_CallMeMacy, and @speedrun among many others. #AI #AgenticEconomy #EnterpriseAI #NYTechWeek #AI #investors – at New York, NY

A few standout moments from #NYTechWeek: ★ @AI_AInstitute’s Chief AI Officer Summit. @AstraZeneca’s pathology computer vision case study showed how regulated-industry leaders design security and governance from day one. ★ @AlphaSenseInc’s Technology Leaders Dinner. Executives from leading financial services and healthcare firms discussed the critical role of data for AI success and scaling transformation. ★ @CarnegieMellon’s VentureBridge Demo Day. At this invitation-only showcase, investors got an exclusive look at deep-tech spinouts from top PhD programs, including @gather_ai ★ @a16z / @speedrun AI Faire. Outstanding interactive showcase of cross-industry, early-stage innovators. —— What were your key takeaways from #NYTechWeek? Especially interested in hearing your insights about scaling AI in highly regulated environments? —— Companies represented included @nvidia, @Databricks, @datadoghq, @BankofAmerica, @Intuit, @ExaAILabs, @HSBC, @IBM, @Adobe, @Atlassian, @Cloudflare, @FireworksAI_HQ, @Google for Startups, @googledevs, @Hyperagent, @fin_ai, @PwC, @vercel, @intercom, @Vanta, @Fiserv, @Citi, @Barclays, @DeutscheBank, @Capgemini, @fgs_global, @Ipsos, and more. #AI #AgenticEconomy #Investors #StartupGrowth – at New York, NY

Only 7% of asset managers have deployed agentic AI — even though 78% are actively exploring it. 88% cite regulatory and compliance as the single biggest barrier. The blocker is not the model. An agent on legacy infrastructure cannot be given bounded authority, an auditable decision trail, or the sign-off a compliance officer needs to let it touch client assets. Autonomy in asset management is an infrastructure problem before it is an intelligence problem. via @EYnews ey.com/en_us/insights/wealth…

Responsible AI: Cost center, or competitive advantage? 🛡️ At the AI Hot 100 Summit with @Mastercard, @FICO_corp, @MorganStanley, @OneCarlyle & @Experian, it was clear: trust, explainability, auditability & continuous monitoring grow adoption, scale, and ROI—not overhead. 📈 Trust is especially critical as agentic systems proliferate & go live in production. Your build-vs-buy decisions across the full AI stack (especially auditable workflows with human-in-the-loop + immutable provenance) determine whether enterprise AI initiatives stall in pilot or deliver real business value. 🚀 How are you turning Responsible AI into ROI? 👇 Full post + framework: bit.ly/4dflx2k _________________ Thanks to the @theaifurnace for the personal VIP invite. #ResponsibleAI #AgenticAI #AIGovernance

Proud to be an alum of @CarnegieMellon, where interdisciplinary excellence and philanthropy have been at the heart of its work since 1900. #AI #Innovation

It’s refreshing to see that companies are increasingly using AI agents—like @GoogleDeepMind’s CodeMender—to increase security. #AI #Innovation #TechTrends

Midsize Enterprise CXOs: gain insight into deploying and adopting 45 emerging technologies across six technology domains. @Gartner_inc #ArtificialIntelligence #AI #Innovation – at New York, NY

𝗪𝗵𝗮𝘁 𝗱𝗶𝗱 @Google 𝗹𝗲𝗮𝗿𝗻 𝗳𝗿𝗼𝗺 𝗮𝘀𝗸𝗶𝗻𝗴 𝟯,𝟱𝟬𝟬 𝗲𝘅𝗲𝗰𝘂𝘁𝗶𝘃𝗲𝘀, “𝗪𝗵𝗲𝗿𝗲 𝗶𝘀 𝗔𝗜 𝗺𝗮𝗸𝗶𝗻𝗴 𝘆𝗼𝘂 𝗺𝗼𝗻𝗲𝘆?” If you’re curious about what’s generating returns in AI right now, check out the slides below. 𝗞𝗘𝗬 𝗜𝗡𝗦𝗜𝗚𝗛𝗧𝗦 1️⃣ Executives are seeing ROI on ≥1 GenAI use cases, especially 88% of early adopters of Agentic AI 2️⃣Data privacy and security are the #1 consideration for companies when evaluating LLM providers 3️⃣Google Cloud’s playbook for deploying AI Agents that deliver positive ROI is on Page 4 #AI #ArtificialIntelligence #Innovation –––––––– 𝘛𝘩𝘢𝘯𝘬𝘴 𝘵𝘰 @neilhoyne 𝘧𝘰𝘳 𝘴𝘩𝘢𝘳𝘪𝘯𝘨 𝘵𝘩𝘦 𝘧𝘶𝘭𝘭 48-𝘱𝘢𝘨𝘦 𝘳𝘦𝘱𝘰𝘳𝘵, 𝘸𝘩𝘪𝘤𝘩 𝘪𝘴 𝘦𝘯𝘵𝘪𝘳𝘦𝘭𝘺 𝘵𝘩𝘦 𝘱𝘳𝘰𝘱𝘦𝘳𝘵𝘺 𝘰𝘧 𝘎𝘰𝘰𝘨𝘭𝘦 𝘢𝘯𝘥 𝘎𝘰𝘰𝘨𝘭𝘦 𝘢𝘭𝘰𝘯𝘦. – at New York, NY

𝗧𝗵𝗲 “𝟵𝟱% 𝗼𝗳 𝗚𝗲𝗻𝗔𝗜 𝗽𝗶𝗹𝗼𝘁𝘀 𝗳𝗮𝗶𝗹” 𝗵𝗲𝗮𝗱𝗹𝗶𝗻𝗲 𝗰𝗼𝗻𝗳𝘂𝘀𝗲𝘀 𝗻𝗼𝗶𝘀𝗲 𝗳𝗼𝗿 𝘀𝗶𝗴𝗻𝗮𝗹. 🏆 𝗟𝗲𝗮𝗱𝗲𝗿𝘀 𝗶𝗻 𝘁𝗵𝗲 𝘁𝗼𝗽 𝟱% 𝘂𝘀𝗲 𝘁𝗵𝗿𝗲𝗲 𝘀𝘁𝗿𝗮𝘁𝗲𝗴𝗶𝗲𝘀 𝘁𝗼 𝘁𝗿𝗮𝗻𝘀𝗳𝗼𝗿𝗺 𝗚𝗲𝗻𝗔𝗜 𝗶𝗻𝘃𝗲𝘀𝘁𝗺𝗲𝗻𝘁𝘀 𝗶𝗻𝘁𝗼 𝗽𝗼𝘀𝗶𝘁𝗶𝘃𝗲 𝗔𝗜 𝗥𝗢𝗜: 1️⃣ REALLOCATE EARLY AI SPEND TO BACK-OFFICE AUTOMATION Executives, on average, allocate 𝟱𝟬% of their GenAI budget to Sales and Marketing according to MIT’s 𝘎𝘦𝘯𝘈𝘐 𝘋𝘪𝘷𝘪𝘥𝘦 2025 study¹. However, back-office process automation yields the clearest near-term ROI.  Target process-specific automations and integrate robust AI solutions—not all of which require GenAI—with existing systems to reduce spend and minimize business process outsourcing. For example, @ATT saved 16.9 million minutes of manual effort per year, recognized “hundreds of millions of dollars in annualized value”, and delivered 𝟮𝟬𝘅 𝗥𝗢𝗜 by scaling an enterprise-wide AI automation program across finance and operations. 2️⃣ LEVERAGE EMPLOYEES’ APPETITE FOR AI EFFICIENCY A thriving “shadow AI economy” exists within enterprises; 𝟵𝟬% of employees surveyed use personal AI subscriptions—such as @OpenAI’s ChatGPT or @AnthropicAI’s Claude—to automate significant portions of their daily work. However, 𝗼𝗻𝗹𝘆 𝟰𝟬% of companies surveyed distribute LLM subscriptions to employees. For AI success at scale, determine which tasks employees already automate, prioritize based on the value added before procuring (or building) new AI tools, and then embed the highest-value workflows into robust, customizable, enterprise-grade tools. 3️⃣ SET REALISTIC TIMELINES FOR ENTERPRISE AI PROJECTS Major technology deployments within multinational organizations typically span 𝟭 𝘁𝗼 𝟯 𝘆𝗲𝗮𝗿𝘀, depending on the scope and depth of integration. The study’s 𝟲-𝗺𝗼𝗻𝘁𝗵 observation window is insufficient to project long-term AI ROI for enterprise-wide deployments. Recalibrate and set organization-wide expectations for the speed of enterprise deployment. _____ Reframe the “95% failure²” statistic as a maturity signal for the 5% that have achieved positive AI ROI, not a stop sign for the 95% yet to realize AI ROI. To maximize enterprise P&L impact and AI ROI, shift early-stage AI spend to back-office automations, analyze shadow-AI usage to identify high-value use cases, and set realistic expectations for enterprise deployment velocity. _____ #ArtificialIntelligence #AIInnovation #AIStrategy ¹ Challapally, Aditya, et al. 𝘛𝘩𝘦 𝘎𝘦𝘯𝘈𝘐 𝘋𝘪𝘷𝘪𝘥𝘦. @medialab Massachusetts Institute of Technology Media Lab, Project NANDA, July 2025. ² Failed AI investments, as defined by the MIT study and confirmed by The @NewYorker, are projects that were not deployed beyond the pilot stage and/or those without measurable return or P&L impact within six months. – at New York, NY

𝗧𝗵𝗲 “𝟵𝟱% 𝗼𝗳 𝗚𝗲𝗻𝗔𝗜 𝗽𝗶𝗹𝗼𝘁𝘀 𝗳𝗮𝗶𝗹” 𝗵𝗲𝗮𝗱𝗹𝗶𝗻𝗲¹ 𝗶𝘀 𝗲𝘃𝗲𝗿𝘆𝘄𝗵𝗲𝗿𝗲. 𝗟𝗲𝗮𝗱𝗲𝗿𝘀 𝗶𝗻 𝘁𝗵𝗲 𝟱% 𝗿𝘂𝗻 𝗮 𝗱𝗶𝗳𝗳𝗲𝗿𝗲𝗻𝘁 𝗽𝗹𝗮𝘆𝗯𝗼𝗼𝗸 𝘁𝗼 𝗰𝗼𝗺𝗽𝗼𝘂𝗻𝗱 𝗔𝗜 𝗥𝗢𝗜. 📈 𝗛𝗲𝗿𝗲’𝘀 𝗵𝗼𝘄, per @MIT’s 𝘛𝘩𝘦 𝘎𝘦𝘯𝘈𝘐 𝘋𝘪𝘷𝘪𝘥𝘦 2025 study: 1️⃣ REALLOCATE EARLY AI SPEND TO BACK-OFFICE AUTOMATION Executives surveyed allocate 50% of GenAI budgets to Sales and Marketing on average; however, back-office automation yields the clearest, near-term ROI. Target process-specific operational automations and integrate robust solutions with existing systems to reduce spend and minimize business process outsourcing. 2️⃣ LEVERAGE EMPLOYEES’ STRONG APPETITE FOR AI EFFICIENCY A thriving “shadow AI economy” exists within enterprises; 90% of employees surveyed use personal AI subscriptions—such as @OpenAI’s ChatGPT or @AnthropicAI’s Claude—to automate significant portions of their daily work. However, only 40% of companies surveyed have enterprise LLM subscriptions. For AI success at scale, determine which tasks employees already automate, prioritize based on the value added before procuring (or building) new AI tools, and then embed the highest-value workflows into robust, customizable, enterprise-grade tools. 3️⃣ SET REALISTIC TIMELINES FOR ENTERPRISE TECHNOLOGY INTEGRATION Major technology deployments within multinational organizations typically span 1 to 3 years, depending on the scope and depth of integration. The study’s 6-month observation window is likely insufficient to project long-term AI ROI for enterprise-wide deployments. Recalibrate org-wide expectations for the speed of enterprise deployment. _____ Reframe the “95% failure” statistic as a maturity signal, not a stop sign. To maximize enterprise P&L impact and AI ROI, shift early-stage AI spend to back-office automations, analyze shadow-AI usage to identify high-value use cases, and set realistic expectations for enterprise deployment velocity. #ArtificialIntelligence #AI #AIStrategy #GenerativeAI #Leadership ¹Original study: Challapally, Aditya, et al. The GenAI Divide: State of AI in Business 2025. @MIT Media Lab, Project NANDA, July 2025. – at New York, NY