TCG

From our CTO, Robert Buccigrossi: 𝐀𝐈 𝐚𝐝𝐨𝐩𝐭𝐢𝐨𝐧 𝐝𝐨𝐞𝐬𝐧'𝐭 𝐟𝐚𝐢𝐥 𝐛𝐞𝐜𝐚𝐮𝐬𝐞 𝐨𝐟 𝐭𝐡𝐞 𝐭𝐨𝐨𝐥𝐬—𝐢𝐭 𝐟𝐚𝐢𝐥𝐬 𝐛𝐞𝐜𝐚𝐮𝐬𝐞 𝐨𝐟 𝐡𝐨𝐰 𝐢𝐭 𝐢𝐬 𝐦𝐚𝐧𝐚𝐠𝐞𝐝. New research from DORA shows that successful AI adoption has far more to do with leadership practices than access to technology. The biggest drivers of adoption aren’t shiny platforms, but clarity, trust, and discipline. What actually works: • Clear AI policies that remove risk and ambiguity. Organizations with clear AI acceptable-use policies saw a 451% 𝐢𝐧𝐜𝐫𝐞𝐚𝐬𝐞. • Dedicated time for teams to learn and experiment. developers dedicated time during work hours to experiment with AI leads to a 131% 𝐢𝐧𝐜𝐫𝐞𝐚𝐬𝐞 𝐢𝐧 𝐭𝐞𝐚𝐦 𝐚𝐝𝐨𝐩𝐭𝐢𝐨n. • Directly addressing job-displacement fears. Organizations that actively address concerns about job displacement see 125% higher adoption. • Actively encouraging AI use in daily workflows (not just coding) • Transparent communication about AI goals and success metrics Just as important: measure impact without creating a culture of surveillance. Increased AI usage only matters if delivery stability improves—not if it erodes. Read the entire blog post here: https://lnkd.in/gYREJdtk If you would like to receive more insights, follow TCG's LinkedIn page. #DigitalTransformation #FederalContracting #ArtificialIntelligence #ResonsibleAI

𝐈𝐬 𝐀𝐈 𝐬𝐩𝐞𝐞𝐝𝐢𝐧𝐠 𝐮𝐩 𝐲𝐨𝐮𝐫 𝐝𝐞𝐯𝐞𝐥𝐨𝐩𝐞𝐫𝐬—𝐨𝐫 𝐪𝐮𝐢𝐞𝐭𝐥𝐲 𝐬𝐥𝐨𝐰𝐢𝐧𝐠 𝐲𝐨𝐮𝐫 𝐝𝐞𝐥𝐢𝐯𝐞𝐫𝐲? 𝘛𝘩𝘦 𝘥𝘢𝘵𝘢 𝘴𝘶𝘨𝘨𝘦𝘴𝘵𝘴 𝘵𝘩𝘦 𝘢𝘯𝘴𝘸𝘦𝘳 𝘥𝘦𝘱𝘦𝘯𝘥𝘴 𝘰𝘯 𝘩𝘰𝘸 𝘺𝘰𝘶 𝘶𝘴𝘦 𝘪𝘵. From our CTO, Robert Buccigrossi. Here’s a follow up to my post from last week. As I mentioned there, the data on AI adoption paints a nuanced picture. It is not a magic switch that instantly doubles velocity. Instead, it offers modest gains in specific areas while introducing new risks to stability if not managed correctly. Specifically, AI should be used to generate small units, not massive walls of code. 𝐓𝐡𝐞 "𝐕𝐚𝐜𝐮𝐮𝐦 𝐇𝐲𝐩𝐨𝐭𝐡𝐞𝐬𝐢𝐬" 𝐚𝐧𝐝 𝐃𝐞𝐥𝐢𝐯𝐞𝐫𝐲 𝐈𝐧𝐬𝐭𝐚𝐛𝐢𝐥𝐢𝐭𝐲 Crucially, DORA’s research indicates that AI adoption can negatively impact delivery performance. They observed a 7.2% reduction in delivery stability and a 1.5% reduction in throughput as AI adoption scaled. This leads to the "Vacuum Hypothesis". AI allows developers to generate code faster. Without discipline, this results in larger batch sizes (larger pull requests). Larger PRs are historically harder to review, slower to merge, and more prone to introducing bugs. Key Takeaway: 𝐌𝐚𝐢𝐧𝐭𝐚𝐢𝐧 𝐬𝐦𝐚𝐥𝐥 𝐛𝐚𝐭𝐜𝐡 𝐬𝐢𝐳𝐞𝐬. AI should be used to iterate faster on smaller units of work, not to generate massive walls of code that overwhelm reviewers. Follow TCG on LinkedIn to get more insights like this! #DigitalTransformation #FederalContracting #ArtificialIntelligence #ResonsibleAI

Last week, the American Council for Technology - Industry Advisory Council (ACT-IAC) held its Topgolf kickoff for the 2026 Associates Program. It was a wonderful opportunity to connect with others in the industry and meet some of the prospective 2026 Associates. Having gone through the program nearly two years ago, I can attest to its significance for both personal and professional development. I highly recommend it to those currently in the industry who are looking to broaden their opportunities for engagement. If you are interested in the Associates Program, you can apply at https://lnkd.in/e7hJVAWQ. Applications are due by midnight EST on February 26th, 2026. For graduates of the program keep an eye out for the upcoming gAp events! #ACTIAC #AssociatesProgram #EmergingLeaders #GovCon #LeadershipDevelopment #gAp Joe Ehrlich Marc Agossou Matt Schmitt,PMP, SAFe Agilist Habib Begham Christopher Goodrich Colin Corey Keara Heck Rebecca Angerome, MBA Robby Singh

𝐃𝐨𝐞𝐬 𝐀𝐈 𝐈𝐧𝐜𝐫𝐞𝐚𝐬𝐞 𝐭𝐡𝐞 𝐕𝐞𝐥𝐨𝐜𝐢𝐭𝐲 𝐨𝐟 𝐃𝐞𝐯𝐞𝐥𝐨𝐩𝐦𝐞𝐧𝐭 𝐓𝐞𝐚𝐦? 𝐇𝐞𝐫𝐞’𝐬 𝐖𝐡𝐚𝐭 𝐭𝐡𝐞 𝐃𝐚𝐭𝐚 𝐒𝐡𝐨𝐰𝐬. From our CTO, Robert Buccigrossi Last month, at QCon NY 2025, Al Crowley and I attended several sessions on effective use of AI in software development from real-world practitioners including Google, NVidia, and Netflix. One presentation stood out for using real-world data for effective AI adoption and management: "Leadership in AI-Assisted Engineering" by Justin Reock of DX. In his presentation he described key lessons in recent research from DORA (Google Cloud’s DevOps Research and Assessment) and DX. 𝘛𝘩𝘦 𝘝𝘰𝘭𝘢𝘵𝘪𝘭𝘪𝘵𝘺 𝘰𝘧 𝘈𝘐 𝘈𝘥𝘰𝘱𝘵𝘪𝘰𝘯 The data on AI adoption paints a nuanced picture. 𝐈𝐭 𝐢𝐬 𝐧𝐨𝐭 𝐚 𝐦𝐚𝐠𝐢𝐜 𝐬𝐰𝐢𝐭𝐜𝐡 𝐭𝐡𝐚𝐭 𝐢𝐧𝐬𝐭𝐚𝐧𝐭𝐥𝐲 𝐝𝐨𝐮𝐛𝐥𝐞𝐬 𝐯𝐞𝐥𝐨𝐜𝐢𝐭𝐲. Instead, it offers modest gains in specific areas while introducing new risks to stability if not managed correctly. Data from over 20,000 developers reveals a "GenAI Divide." Averages hide extreme volatility. For every organization realizing a massive gain in change confidence, another sees a commensurate drop. Success depends entirely on implementation strategy, not just tool access. 𝘖𝘯 𝘢𝘷𝘦𝘳𝘢𝘨𝘦, 𝘢 25% 𝘪𝘯𝘤𝘳𝘦𝘢𝘴𝘦 𝘪𝘯 𝘈𝘐 𝘢𝘥𝘰𝘱𝘵𝘪𝘰𝘯 𝘤𝘰𝘳𝘳𝘦𝘭𝘢𝘵𝘦𝘴 𝘸𝘪𝘵𝘩: • 7.5% increase in documentation quality. • 3.4% increase in code quality. • 3.1% increase in code review speed. In the context of this data, I wrote about keys to effective AI adoption for TCG's blog. If you're interested, take a look! No sign up required :) https://lnkd.in/gYREJdtk

From our CTO, Robert Buccigrossi! What happens when your people adopt AI faster than your policies do? The rush to adopt Generative AI has created a massive blind spot inside most organizations — one we rarely acknowledge. At a recent CSA webinar, security researchers from Layer X laid out the reality: most enterprise AI usage is happening far outside official channels, with almost no oversight. Here are the biggest takeaways every technical leader should understand: 1. Enterprise AI is a Shadow Ecosystem Policies don’t reflect real usage. Research shows 89% of enterprise AI activity is invisible to security teams, and 71% of traffic comes from personal accounts — many not protected by SSO. That data? It often lands in consumer AI systems, gets stored indefinitely, and leaves no audit trail. 2. Data Leakage Happens in Bulk Security policies often worry about “prompts,” but the real exposure happens when users paste: -Source code -HR and finance documents -PII or customer data Samsung learned this the hard way when engineers pasted confidential code into ChatGPT — forcing an internal ban and emergency policy rewrite. 3. Browser AI Extensions Are Silent Threats One in five enterprise users has an AI browser extension installed — 60% run with high-privilege access. These plugins can read session tokens, capture credentials, and exfiltrate data without triggering traditional endpoint controls. 4. Banning AI Doesn’t Work 27% of organizations have tried full prohibition. But blocklists just create Shadow AI. Employees still turn to GenAI — they just move to personal devices, VPNs, and unmonitored apps. 5. The Browser Is the Real Control Point AI tools, embedded AI inside SaaS, and risky extensions all meet in one place: the browser. Security teams need real-time visibility at the point of interaction, including: - enforcing corporate accounts - educating users in session  -blocking copy/paste or upload of sensitive data  -monitoring long-tail AI tools Out-of-band controls simply can’t keep up. The Only Sustainable Path Organizations must enable AI safely, not pretend usage isn’t occurring. AI accelerates mission delivery — but only when we use it responsibly. Let’s stay curious, stay smart, and throw some light on the shadows.

Working at TCG, I’ve learned a great deal about how thoughtful engineering and architecture shapes long-term outcomes, and one of the takeaways that I have noticed is technical debt. Technical debt is often framed as messy code, but its real impact is architectural. Tight coupling between components, undocumented integrations, and outdated patterns create systems that resist change. Over time, even small updates require excessive testing, manual workarounds, and increased risk. This debt directly affects security, scalability, and delivery velocity. When architecture cannot evolve, teams compensate with process instead of improvement. In regulated environments, architectural debt also slows compliance, as controls are bolted onto systems never designed to support them. Addressing technical debt requires intentional refactoring, clear system boundaries, and governance that prioritizes long-term sustainability over short-term delivery pressure. #GovTech #DigitalTransformation #TCG

TCG is honored to be ranked as one of the 2026 #TopImpactCompanies by @Real Leaders magazine! This year’s honorees were chosen for their measurable impact—innovators transforming their industries and redefining what it means to lead with purpose. https://lnkd.in/ejmV8ftX

Thank you GovCIO for hosting the Zero Trust Forum. As directed by the White House, EO 14028 “Improving the Nation’s Cybersecurity” requires federal agencies to implement Zero Trust Architecture (ZTA) principles for enhanced security. I took the opportunity to attend the Federal Zero Trust Forum last week where I was able to gain additional insight on the strategic approach, expected workforce and cultural challenges, and lessons learned in integrating ZTA within government systems. It was a great experience to come together with tech leaders across the federal government and learn more about building a more secure government. TCG GovCIO Media & Research

Thrilled to share some reflections as we close out the year. One of the biggest lessons I’ve learned in the government contracting space is that adaptability isn’t just a skill — it’s a requirement. Whether it’s shifting priorities, evolving compliance needs, or responding to changes impacting federal agencies, the teams that stay proactive always rise above the challenge. Government contracting is never static — policies shift, budgets tighten, operational needs evolve. But one thing that remains constant is the importance of reliable, mission-driven partners. At TCG, I’ve seen firsthand what it looks like when a team stays committed to clarity, communication, and client success even during uncertain times. I’m grateful to work alongside talented people at TCG who bring excellence, accountability, and heart into every project. Looking forward to continuing to support our clients, strengthen operations, and embrace new opportunities for growth heading into the new year. #TCG #WorkWithPurpose

From TCG's CTO Robert Buccigrossi: A client application was flagged as a "Cryptocurrency Wallet Drainer." It wasn't. Here's what happened. #GovTech #FederalIT #FederalCybersecurity #cybersecurity https://lnkd.in/eiCPnH3a