Even as a teenager, Javier Rando was fascinated by technology,\u00a0computers\u00a0and the idea that robots could one day take over human tasks.\u00a0The\u00a026 year-old\u00a0native of\u00a0M\u00e1laga, Spain, was\u00a0never\u00a0a gamer, but\u00a0he did programme\u00a0a smartphone app on his computer\u00a0before he\u00a0even\u00a0had\u00a0his own\u00a0smartphone. Later, his fascination with technology\u00a0took\u00a0him to Barcelona to study mathematical engineering and data science. It was there that he first\u00a0encountered\u00a0artificial intelligence, particularly through the application of algorithms and machine learning.\u00a0<\/p>\n
It was during his\u00a0Bachelor\u2019s degree\u00a0that\u00a0his image of artificial intelligence\u00a0first\u00a0began to\u00a0falter. \u201cI was horrified,\u201d he recalls. He had read in the media that courts in the United States were using AI to assess the risk of\u00a0recidivism\u00a0among offenders. However, this AI systematically classifies people as more dangerous solely because\u00a0they are\u00a0Black. \u201cThat changed my focus. I decided to devote my\u00a0Bachelor thesis\u00a0to fairness in artificial intelligence,\u201d says Rando.\u00a0<\/p>\n
Enormous potential \u2013 both positive and negative <\/p>\n
After that he got\u00a0more and more\u00a0interested in the topic of AI safety.\u00a0He followed up his\u00a0Bachelor\u2019s\u00a0with a\u00a0Master\u2019s degree in\u00a0Computer\u00a0Science\u00a0at ETH Zurich. \u201cI chose ETH because I wanted to study at the best university for computer science in Europe,\u201d he says.\u00a0He was fascinated by\u00a0Professor Florian\u00a0Tram\u00e8r’s\u00a0research on AI safety,\u00a0so\u00a0he applied for a PhD. He was accepted\u00a0and also\u00a0received a fellowship position at the ETH AI\u00a0Center.\u00a0<\/p>\n
AI\u00a0Center\u00a0Fellows typically work\u00a0across disciplines, combining different areas of AI. Rando focused on the safety of language models. \u201cToday, language models are the most widely used form of artificial intelligence.\u00a0Millions of people use them. This means they also pose the greatest risk of causing harm,\u201d he\u00a0says. At the same time, he emphasises that his view of AI is not fundamentally negative: \u201cArtificial intelligence has the potential to become one of the best technologies in human history.\u201d It could simplify our lives,\u00a0such as\u00a0by developing\u00a0cures\u00a0for diseases and much more. But there are many risks along the way.<\/p>\n
PhD studies in San Francisco <\/p>\n
AI is not yet so advanced that its dangers are\u00a0particularly\u00a0great. \u201cBut that will change. We are on a dangerous path because\u00a0the\u00a0progress is\u00a0lightning fast\u00a0and we\u2019re developing\u00a0a very powerful\u00a0technology,\u201d says Rando.\u00a0Most people are unaware of this because they use AI for harmless purposes. Anyone who asked\u00a0it\u00a0for a recipe two years ago and today will not notice\u00a0a huge difference.\u00a0But\u00a0that belies the potential of AI.\u00a0The risks increase significantly as soon as people begin\u00a0using\u00a0AI\u00a0more\u00a0and more\u00a0as so-called agents.\u00a0Examples of this include\u00a0instructing\u00a0an\u00a0AI to respond to emails automatically.\u00a0This will be of interest to criminals who\u00a0can\u00a0attempt\u00a0to email the AI\u00a0a command\u00a0to disregard the task set by the computer user and\u00a0to\u00a0send\u00a0them\u00a0their credit card details instead.<\/p>\n
Rando sees two fundamental dangers: first, AI is vulnerable to attacks and manipulation. \u201cThere\u2019s\u00a0still a long way to go before these\u00a0loopholes\u00a0are closed,\u201d he says. Second, AI can be misused by people with bad intentions even without manipulation\u00a0\u2013 it could even\u00a0create instructions for building weapons.\u00a0<\/p>\n
\u201cBut the\u00a0benefits and\u00a0opportunities outweigh the risks. That\u2019s why I\u2019m researching ways to make artificial intelligence safer,\u201d says Rando. He is working on this in his doctoral studies and in various professional roles, which have taken him to high-profile AI companies such as OpenAI, Meta and now Anthropic.\u00a0After spending the first part of his doctoral studies in Zurich, he has been living in San Francisco since early 2025 as an external doctoral\u00a0candidate, while also\u00a0working\u00a0at\u00a0US\u00a0AI company Anthropic, the manufacturer of the AI \u201cClaude\u201d which is subject to strict ethical guidelines.<\/p>\n
Searching for vulnerabilities <\/p>\n
Both in his doctoral studies and at Anthropic, Javier\u00a0Rando\u00a0is working on simulating manipulations and attacks on artificial intelligence, searching for\u00a0vulnerabilities\u00a0and closing these gaps with\u00a0appropriate programming. At the same time, the aim is to prevent people with malicious intentions from being able to misuse AI in the first place. \u201cWe need to build protective shields around AI,\u201d he says.\u00a0<\/p>\n
The aim is for artificial intelligence to understand when humans are asking it to perform harmful actions and to be able to refuse to carry out such tasks. This requires\u00a0appropriate programming, but not all AI manufacturers consider this to be equally important: \u201cIn AI development, the bottom line has to be safety first, not profit.\u201d That\u00a0calls for a regulatory\u00a0framework. Javier recently contributed his\u00a0expertise\u00a0as an advisor to a European Commission working group\u00a0on\u00a0developing guidelines for AI manufacturers, which are voluntary for the time being.<\/p>\n
Javier Rando\u2019s work at Anthropic and his research at ETH Zurich are closely intertwined. His dissertation will consist of several publications, for which he can also\u00a0rely on\u00a0the research\u00a0he has\u00a0conducted at Anthropic.\u00a0When asked about the difference between researching\u00a0at universities and\u00a0at\u00a0tech companies, he says,\u00a0\u201cBoth approaches are needed to make the AI of the future sufficiently safe and to be able to reap its\u00a0rewards.\u201d\u00a0The advantage of being\u00a0a researcher\u00a0at\u00a0a company is\u00a0that it gives you\u00a0access to unlimited computing power and\u00a0you\u00a0know the company\u2019s AI inside out.\u00a0While the upside of researching at a university\u00a0is that it enables you\u00a0to explore and shape research questions more freely and to test solutions that are\u00a0more risky\u00a0from an economic perspective.\u00a0<\/p>\n
Over the next few years, Javier Rando plans to continue researching the\u00a0safety\u00a0of AI applications at Anthropic while completing his\u00a0PhD.\u00a0And after that?\u00a0It\u2019s\u00a0still unclear what the future holds. \u201cDevelopments in the field of AI are happening so\u00a0fast\u00a0that we can\u2019t predict what the world will look like in two years\u2019\u00a0time.\u00a0Perhaps in\u00a0a few years\u2019\u00a0time,\u00a0I\u2019ll\u00a0be replaced by AI\u00a0at work as well,\u201d\u00a0says\u00a0Rando.\u00a0<\/p>\n","protected":false},"excerpt":{"rendered":"Even as a teenager, Javier Rando was fascinated by technology,\u00a0computers\u00a0and the idea that robots could one day take…\n","protected":false},"author":2,"featured_media":71512,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":"","_share_on_mastodon":"0"},"categories":[6],"tags":[980,8614,272,1202,25339,51],"class_list":["post-71511","post","type-post","status-publish","format-standard","has-post-thumbnail","category-zurich","tag-careers","tag-computer-and-information-technology","tag-international","tag-news","tag-security-research","tag-zurich"],"share_on_mastodon":{"url":"https:\/\/pubeurope.com\/@ch\/116634023723454419","error":""},"_links":{"self":[{"href":"https:\/\/www.europesays.com\/ch\/wp-json\/wp\/v2\/posts\/71511","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.europesays.com\/ch\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.europesays.com\/ch\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/ch\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/ch\/wp-json\/wp\/v2\/comments?post=71511"}],"version-history":[{"count":0,"href":"https:\/\/www.europesays.com\/ch\/wp-json\/wp\/v2\/posts\/71511\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/ch\/wp-json\/wp\/v2\/media\/71512"}],"wp:attachment":[{"href":"https:\/\/www.europesays.com\/ch\/wp-json\/wp\/v2\/media?parent=71511"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.europesays.com\/ch\/wp-json\/wp\/v2\/categories?post=71511"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.europesays.com\/ch\/wp-json\/wp\/v2\/tags?post=71511"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}