CEO LinkedIn cảnh báo: AI sẽ gây hỗn loạn nghề nghiệp – kỹ năng con người mới là vũ khí sống còn

• Ryan Roslansky, CEO LinkedIn, cho biết AI sẽ mang lại sự gián đoạn lớn trên thị trường lao động và không dễ dàng như nhiều người lầm tưởng.

• Các công ty như Amazon cũng xác nhận đang cắt giảm nhân sự do AI giúp tự động hóa nhiều tác vụ nội bộ.

• Roslansky nhấn mạnh: “Không thể làm ngơ. Phải chủ động thích nghi.” Và chìa khóa chính là phát huy kỹ năng mang tính con người như giao tiếp, cộng tác và tư duy phản biện.

• Trên LinkedIn, số lượng kỹ năng liên quan đến AI trong các tin tuyển dụng tăng gấp 6 lần trong 1 năm, trong khi số người thêm kỹ năng AI vào hồ sơ tăng gấp 20 lần.

• LinkedIn triển khai tính năng AI hỗ trợ tạo hồ sơ và tìm việc dễ dàng hơn, thay vì người dùng phải tìm kiếm bằng từ khóa như trước.

• Dù vậy, Roslansky cho biết người dùng LinkedIn rất cẩn trọng khi dùng AI viết bài, vì đây là nơi xây dựng uy tín chuyên nghiệp. Việc bị phát hiện "viết bằng AI" có thể làm giảm cơ hội nghề nghiệp.

• Với vấn nạn tài khoản giả do AI tạo ra, LinkedIn khuyến khích người dùng xác minh danh tính qua email công ty, giấy tờ cá nhân… để tạo niềm tin trong mạng lưới.

• Roslansky tin rằng AI sẽ tạo ra nhiều việc làm mới, đặc biệt là trong lĩnh vực khởi nghiệp, sáng tạo nội dung, xây dựng ứng dụng và học tập phi truyền thống.

• Tương lai vẫn sáng nếu biết tái đào tạo kỹ năng và học hỏi liên tục, dù ở vai trò kỹ thuật hay phi kỹ thuật.

• Roslansky chia sẻ cá nhân rằng ông thường dùng Copilot để viết email gửi CEO Microsoft Satya Nadella, và AI đã biết cả lịch sử tương tác giữa hai người — một ví dụ cho thấy AI cá nhân hóa đang ngày càng sâu sắc.

📌 CEO LinkedIn Ryan Roslansky dự báo AI sẽ gây ra làn sóng gián đoạn lớn, nhưng “kỹ năng con người” như giao tiếp, tư duy và hợp tác vẫn là công cụ phân biệt cốt lõi. Số kỹ năng AI trên LinkedIn tăng gấp 20 lần, nhưng người dùng vẫn cần giữ chất riêng để nổi bật giữa làn sóng tự động hóa. Chuyển đổi nghề nghiệp là tất yếu, và học suốt đời là bắt buộc.

https://www.bloomberg.com/news/newsletters/2025-06-20/linkedin-s-ceo-has-advice-to-help-ai-proof-your-career?srnd=technology-ai

June 20, 2025 at 7:00 PM UTC

LinkedIn’s CEO offers his thoughts on how AI is changing the way people work, himself included. But first…

Three things to know:

• Masa Son pitches $1 trillion US AI hub to TSMC, Trump team
• OpenAI is phasing out Scale AI work following startup’s Meta deal
• Musk’s xAI burns through $1 billion a month as costs pile up

Standing out in an AI world

Ryan Roslansky has many reasons to be optimistic about AI. As the chief executive officer of LinkedIn and a key leader at OpenAI backer Microsoft, Roslansky and his team have had access to generative AI features longer than most. He’s seen how AI can surface relevant jobs for people more easily on the social platform and improve daily workflow.

Like many in Silicon Valley, Roslansky thinks AI breakthroughs may eventually help solve intractable problems like climate change and poverty, as well as enable a new era of “democratization” where anyone can build an application or start a small business. But Roslansky also cautions that there’s discomfort on the horizon for workers.

“It’s Pollyanna to believe that this transition just easily occurs,” Roslansky told me in an interview in late May. “There’s going to be a ton of disruption. There’s going to be a ton of uncertainty along the way.”

Some of that uncertainty is already here. Amazon CEO Andy Jassy said this week that he expects the company’s workforce to decline in the next few years as AI is used to handle more tasks. Other companies have also recently said AI is causing them to rethink headcount.

To weather the AI shift, Roslansky said workers must first avoid burying their heads in the sand. “You can’t just say, ‘This isn’t a thing.’ You have to really jump into it and adapt.”

 

What LinkedIn’s data shows, he said, is that human skills remain a valuable differentiator. “Whatever is uniquely human about yourself, lean into that. Communication, collaboration, all those things, be really good at that,” he said. “That could be the thing that actually helps you stand out.”

Roslansky took over as LinkedIn CEO five years ago, helping to expand its annual revenue from $7 billion to $17 billion. Recently, he was also tasked with overseeing the teams behind the Microsoft Office bundle.

In our conversation, we spoke about the areas where AI has and has not taken off among LinkedIn’s users, as well as the platform’s challenges around the onslaught of fake AI-generated accounts. This interview has been edited for length and clarity.

Bloomberg: It was recently your five-year anniversary as CEO of LinkedIn. Congrats! What are you most proud of and what has been the biggest challenge in those five years?

Roslansky: There’s a lot of really good things that are happening on this platform. If I was to say what I’m most proud of, though — managing a social platform is a really unique opportunity. Keeping the community safe, trusted and professional, especially through Covid, through a lot of change in the world, has been a really important focus of ours. We’re lucky that we have five different business models that run through LinkedIn. We didn’t chase engagement for engagement’s sake. Looking back over the past five years, we’ve been able to keep LinkedIn as that safe, trusted professional network.

AI has been around for a while and LinkedIn had been using it before ChatGPT. How has this new wave of generative AI changed the way that you’re running the business, from staffing to products to how users are engaging with it?

Generative AI has impacted all three of the things you’re talking about. First and foremost, the member base. There’s roughly 15 million jobs posted on LinkedIn at any given time. Over the last year, there’s been a 6x increase in the skills required in any of those jobs being AI-related. On the flip side, over the past year, members adding an AI skill to their profile is up 20x. We’re seeing this AI wave come across the member base.

 

On the way the actual product works — that’s probably been our biggest focus. A lot of people join LinkedIn and it’s a scary experience. It’s their resume online and they’re like, “Oh my gosh, what do I say about myself?” You can use AI to help portray yourself in the right way, or make sure you talk about these skills in a certain way. Helping people establish their identity with AI was one of the most important things that we started with.

 

Then it got a little more sophisticated: Leveraging generative AI for the way that a lot of our existing AI models work, so the feed is being run basically as a large prompt. Here’s Shirin. Here’s what she’s done in the past. Here’s her profile. What should she see?

And we just launched this AI job search. Historically, it’s a classified model that’s existed for 20 years. You have to know which keywords to search for and choose a bunch of facets to try and narrow the jobs down. What AI allows us to do is basically just say what you want to do or what you’re trying to do.

Everyone’s leaning into AI agents hard right now, but there’s varying degrees of how well they can execute on tasks. How effective are you finding agents for tasks like recruiting?

To be upfront about it, I would say probably 18 months ago, it wasn’t great. With all of this stuff, you can’t just make this assumption that it’s going to be out-of-the-box great. Fine-tuning a lot of these models on top of LinkedIn data, and the things that we’re learning from beta customers using it and providing feedback — that’s the key to all of this.

The basic task of, “Hey, there’s a billion people on LinkedIn. I need to find someone like [them]” — it’s really good at that. Step two: “Reach out to this person to make sure that she’s actually in the market for a job or wants to have a conversation.” Historically, it’s not great at that. But that’s a pretty low-level task we’ve gotten really, really good at. “Talk to this person and convince them to come work at LinkedIn?” No way. A human being is still way better at that sales process. That’s where I think that the tools stop. I don’t know if they’ll ever really catch up.

 

How much are people using AI to help them write posts on LinkedIn?

We have an ability where you write a post on LinkedIn. We are not going to write it for you from scratch, but if you’re like, “Hey, help me make this sound better,” you can push a button and we can give you some suggestions. It’s not as popular as I thought it would be, quite frankly.

Why is that?

Taking a step back, this is your resume online. This is your professional reputation in general, which means that people are less likely to share anything on LinkedIn than they are somewhere else because the barrier is much higher. Secondly, we want to make sure that they sound really good when they do it.

Oftentimes, when something comes across as being very AI-obvious on the platform, the rest of the community will call you out. If you’re getting called out on X or TikTok, that’s one thing. But when you’re getting called out on LinkedIn, it really impacts your ability to create economic opportunity for yourself. People are watching the community and if they are using AI tools, they’re going back and making sure that it feels like there’s a human, authentic touch on top of it.

How do you deal with how easy it is now to create a fake LinkedIn profile because of AI?

Historically, it was great because we didn’t really have to worry about this. Creating a fake profile wasn’t easy and it wasn’t worth it. Now it’s easy. We obviously use AI to try and detect fake profiles when they’re created — and there’s a lot of obvious traits with a fake profile. But the right strategy for us is we allow you to verify yourself on LinkedIn. We have a lot of third-party partners. You can upload a photo ID. You can upload your passport. You can verify yourself with your work email address. When you do that, we put this little check next to your profile. A little check follows you around LinkedIn. It’s not like a badge you can buy, like on other networks.

What are the promising areas where you’re seeing AI actually create new types of jobs? How are you thinking about that?

 

In the long term, it’s a really good thing. I think we’re going to see real democratization in terms of how people can now easily start small businesses, build an app, teach themselves something that used to only be available if you were in some elite educational institution. I think that, in and of itself, is going to create new roles.

It’s Pollyanna to believe that this transition just easily occurs. No. There’s going to be a ton of disruption. There’s going to be a ton of uncertainty along the way. People are going to have to learn how to re-skill now that their job is changing on them, even if they’re not changing their job. Just like there has been in any historical labor market paradigm shift, things usually end up in a much better, amazing place with the evolution of these tools. But there’s this super messy middle, and I think that that is going to be the case with AI. It’s naive to think otherwise.

My hope is that what previous labor shifts have lacked, which exists now, is LinkedIn to help make this an easier transition by sharing this data, help people connect, learn things, share knowledge with each other and have access to the right opportunity.

We did see layoffs at Microsoft impact coding jobs disproportionately. Did LinkedIn see cuts in those areas? And would you attribute any of that to AI?

 

Nothing’s attributed to AI. But like every large company right now, we need to make sure that we are on the bleeding edge of ensuring that everyone in our company is becoming more AI-first in the way that they’re thinking about doing their jobs. If it’s in recruiting, using our recruiting tools. If you’re an engineer, it’s really leveraging GitHub Copilot. It’s interesting to watch how a lot of companies take different paths to help their company become an AI-first company. I don’t think you can top-down this. Quite frankly, I think it’s much more of a bottom-up approach.

 

How do you use generative AI, personally?

My boss is Satya Nadella. Every time, before I send him an email, I hit the Copilot button to make sure that I sound Satya-smart. I definitely use it a lot in creating content. And then...

I just have to pause. Do you say, “What would Microsoft CEO Satya Nadella think about there?” Did you cater the AI to his personality?

Yeah, sometimes. Copilot can tell who I’m sending this email to. It has the history. It has access to all of our emails historically. It knows what he’s working on, knows what I’m working on. It’s pretty impressive.

Got a question about AI? Email me, Shirin Ghaffary, and I’ll try to answer yours in a future edition of this newsletter.

Human quote of the week

“We expect that this will reduce our total corporate workforce as we get efficiency gains from using AI extensively across the company.”

Andy Jassy

Amazon CEO

In an email to employees on Tuesday, Jassy said generative AI and AI-powered software agents “should change the way our work is done.” In the coming years, Jassy expects Amazon “will need fewer people doing some of the jobs that are being done today, and more people doing other types of jobs.” He added: “It’s hard to know exactly where this nets out over time.”