Dreamforce 2026: A question of trust

A spectre of trust haunted last week’s Dreamforce conference in San Francisco. The supplier behind the conference, Salesforce, has long vaunted its trustworthiness as a software-as-a-service provider. Its thesis is that as a cloud supplier, trusted with its customers’ data and applications, Salesforce has had to put a premium on trust from its inception. It is in its DNA, and it’s a pioneer of software as a service, founded in 1999, when the concept seemed borderline eccentric. In a press session entitled “The Trust Imperative”, Lindsey Finch, executive vice-president of global privacy, product, artificial intelligence (AI) and cyber security legal, noted that when she started at the company in 2008, “cloud computing was not used as a concept as yet. All the customer conversations, all the policy maker and regulator conversations came down to: ‘you want me to hand over all my data to you?’ That seems somewhat quaint now, in 2026, now that cloud computing is the industry standard. Now, we are in another pivot from an innovation standpoint where trust is at the centre. “We have gone from AI as a chatbot to AI being able to take action,” she said. “That is exciting from an innovation viewpoint, but it does raise the bar in terms of what is necessary from a trust standpoint.” It is the so-called frontier model AI companies, like OpenAI, Anthropic and the rest, with whom Salesforce now readily consorts who have, maybe unwittingly, put trust back on the agenda of enterprise IT, as it was in the early days of cloud computing. Putting it colloquially, “how can we trust those guys?”, described appositely by The Financial Times ’ chief political commentator as “overgrown children”, in an episode of the Political fix podcast. The immediate background to Dreamforce 2026 was the controversy unleashed by the testimony of a young whistleblower from Anthropic, Jacob Coxon , who said AI developers are ever more fearful that the technology they are making has a 10% chance of killing humanity in the next decade. And outside the conference, a group of protestors bewailed Salesforce’s ties to US Immigration and Customs Enforcement (ICE), calling on the company to sign a democracy pledge opposing efforts to block elected candidates, restrict voting and prevent eligible votes from being counted in the nation’s elections. Marc Benioff put the vexed question of trust to Sam Altman, CEO of OpenAI, in a fireside chat at Dreamforce. Altman said he was “very confident in our company’s ability, our industry’s ability, to do this [AI development] safely to make sure that we keep alignment and safety”, but that he was “disappointed about how [the discussion] has been framed”. Benioff asked him to say more about that disappointment, and Altman said: “The world should trust that we are going to do the right thing because it’s the right thing and because we feel the magnitude.” Bahrololoumi and Niles: salience of trust Applying a UK lens to the AI and trust discussion, Zahra Bahrololoumi , CEO of Salesforce UK and Ireland, told a small group of journalists in a briefing at the event that the theme of trust was the main takeaway, for her, from Benioff’s opening conference. “Personally, I felt that the theme of trust and ethical AI and responsible AI, I genuinely felt that came through loud and clear,” she said. “And I think we always wanted to make sure that was front and centre. “When we think about some of the product announcements and the evolution of the platform, the theme of trust, trusted applications, the semantic layer with the Customer 360 and then trusted agents, well, you get the point.” From a Salesforce global corporate point of view, chief legal officer Sabastian Niles went further into the nuances of the concept of trust in an interview with Computer Weekly. Prior to joining Salesforce in 2023, Niles was a partner at New York law firm Wachtell, Lipton, Rosen & Katz. There, he said he had the “privilege” of being a trusted advisor. “But, particularly at Salesforce, our number one value is trust; we take seriously the concept of a profession underlying which is responsibility of obligation,” he said. “All these topics of trust, risk, responsibility, accountability, ethics, are really coming to the fore now. And our customers in industries and governments around the world, they look to us as a trusted advisor, and that’s really powerful.” Read more about trust in data in the age of AI Managing trust in the age of AI influence : can AI technology foster trust and serve the common good, or will algorithms shape humanity’s future for the worse? Experts at EmTech MIT wrestled with the nuances. Data protection and consumer trust are the key to unlocking AI , says Miriam Everett, partner Herbert Smith Freehills. The UK’s National Data Strategy must evolve to unlock AI in the public sector : the UK government is investing both hope and money into using AI to improve public services - but the benefits will remain elusive unless the public sector changes its approach to data strategy, says Matt Harris, managing director for UK, Ireland, Middle East and Africa at HPE. Niles said Salesforce guides itself through all the velocity and variability of the IT industry by grounding itself in its customers’ businesses first. “If your North Star is the ultimate impact, success and effectiveness of the customer, that enables a lot of alignment around trust, and around innovation,” he added. “I’ve never held the view that trust and innovation are at opposite ends of a spectrum.” Nevertheless, Salesforce is a well-established enterprise software company, and in terms of AI, it is partnering with Silicon Valley companies that burst on to the scene in 2022 with large language model technology that is, in essence, autonomous, whatever the conventional talk about “guard rails” or, recently, “ pacing the frontier ”. How does Niles think this tiger can be tamed? “I’ve always firmly believed that the greatest risk is the unexamined status quo,” he said. “That is a very powerful kind of risk. And that also includes how we tackle responsibility, accountability, trust and the like at a time of velocity, uncertainty, ambiguity, volatility. And when you look at all those items, you can see a downside, but I also think there’s all this opportunity. “I love my regulated industries – banks, broader financial services, insurance, healthcare, life sciences,” said Niles. “These are industries that really grapple with leading with technology, trusted platforms and AI agents in a context where responsibility is top of mind … These are industries that historically have been viewed as very cautious, but increasingly, you have a lot of folks in those industries saying ‘OK, there’s a lot of change happening, how do we, highly stakeholder-oriented firms define and lead that change rather have it just happen to us?’ And these are also fascinating industries where they’ve got very deep data, complex workflows, broader obligations, and so on. “Now, if you put those kinds of customers at the centre, put those stakeholders at the centre, that guides you forward to achieve responsible velocity.” If he is right, AI might yet prove to be as unremarkable as cloud computing in the enterprise.