SutherlandGold

Customer service is entering a new era. With @Zendesk acquiring SG client Forethought, its largest acquisition yet, the shift from managing support conversations to AI resolving them autonomously is accelerating. The companies that win will be the ones to adopt AI and use it to expand how far customer support can reach. We’re proud to be a part of Forethought’s journey and look forward to what’s ahead. More here: techcrunch.com/2026/03/11/zen…

AI and cloud are changing virtualization. As AI workloads continue to rise in popularity, they’re creating both new opportunities and new risks across virtualized tech stacks. In IT Brew, reporter Eoin Higgins features insights from SG client Kevin Paige, CISO at @ConductorOneInc, on how enterprises should be thinking about virtualization in the age of AI. A great read if you’re watching how AI is reshaping infrastructure and security. itbrew.com/resources/ai-a…

It likely means rethinking where visibility is built — and how brands show up when audiences search for answers. Key insights👇 & more information in @NiemanLab: bit.ly/4brp4ee

For marketing and communications teams, this raises a bigger question: How do brands stay visible when traditional search-driven traffic becomes less reliable?

Traffic to the top ten English-language tech publications has declined 58% since 2024, according to analysis from @growtika. The way people discover information online is changing. 🧵

Why are some AI startups staying intentionally lean? Why did the NFL choose Bad Bunny for the Super Bowl halftime show? And why are investors prioritizing retention over hype? We unpack it in our latest edition of #BetweentheLines: linkedin.com/pulse/between-…

🔗 in @DarkReading: darkreading.com/application-se… 🔗 in @business: bloomberg.com/news/articles/…

Your AI assistant doesn't need to be hacked. It just needs to be asked to summarize a webpage. Our clients at @hiddenlayersec just published research on @openclaw— 100k+ GitHub stars, one of the fastest-growing open-source AI projects ever built — showing exactly what a real attack looks like from the inside. It starts there. One user. One webpage. No exploit, no elevated privileges. From there, the attack silently executes code, exfiltrates every API key and credential on the system, rewrites the agent's memory so the behavior carries into every future session, and establishes a command-and-control server that phones home every 30 minutes. The user never knows that anything happened. What makes this research worth reading isn't any single vulnerability. It's that none of it requires a sophisticated attacker. It emerges naturally when an AI has persistent memory, tool access, and external communication — and is left to make its own security decisions. That's not a bug in OpenClaw. It's a design pattern showing up across the industry. No sophisticated attacker required. This is what agentic AI looks like when security is left to the model itself.

More from Alex in CRN: crn.com/news/security/…

Autonomous AI attacks are moving faster than most security teams can respond. In a new CRN piece, @AlexBovee, CEO of @ConductorOneInc explains why identity is becoming the first real stress point as automated threats scale. Most attacks already start with identity. AI just speeds up the exploitation. What breaks first: 🔹 MFA gaps 🔹 Over-permissioned accounts 🔹 Slow, ticket-based fixes 🔹 Defenses built for human-speed threats

AI adoption is exploding. ROI? …not so much. @forethought_ai co-founder @dojideon joined @CNBC’s #ForttKnox with @jonfortt, to break down why teams are missing ROI and how agentic AI is finally changing that equation.

Trump’s AI executive order doesn’t clear up the uncertainty — it just shifts it. More in #BetweentheLines: bit.ly/4rYOAgg

AI isn’t just driving model innovation...it’s starting to squeeze the hardware market. ⚡ RAM prices have more than doubled 💻 PC prices projected to rise 23% 🏭 Suppliers prioritizing AI data centers over consumer devices As companies race to build out AI infrastructure, they’re pulling massive volumes of memory into data centers, and everyone else is competing for what’s left. In @nytimes, SG client @M12vc’s Managing Partner Michael Stewart digs into what this means for the global memory market — and why this feels more structural than cyclical: nyti.ms/4cduo5Z

As AI enters real operational use, the big questions are becoming clear: Where is adoption real? What is actually working? What matters more than speed? On @BloombergTV, Cheryl Cheng of @M12vc digs into why trust and operability now matter more than speed or accuracy for real AI deployment.

Most “agent problems” aren’t agent problems. As @Forethought_AI's Antoine Nasr told @CIOonline, early deployment just exposes the routing, language, data, and ownership gaps that were already there. That visibility is the point. bit.ly/4rEJPZj

The reaction to @moltbook showed how quickly a narrative takes shape. A few agent posts appeared and almost immediately turned into talk of breaches, infiltration, “AI societies,” and several competing futures... all drawn from the same screenshots. People filled in the gaps on the spot. That’s the environment now: new tech lands, and meaning forms before anyone has a chance to describe what they’ve actually built. For anyone working in this space, timing matters. If you don’t offer context early, the conversation will create its own version and move on. Moltbook will almost certainly fade. But the instinct to turn early signals into full stories is here to stay.

The most important signals of 2026 are the quiet ones. Our January edition of #BetweentheLines digs into micro-communities, the @goldenglobes' attempt to classify podcasting, @drpepper's timely restraint, and @forethought_ai's focus on measurable AI impact. linkedin.com/feed/update/ur…