BY: ADAM RADOLIŃSKI | FOCUSEUROPE.PL
TRANSLATION: ALEKSANDRA KRZYSZTOSZEK
Media in the Age of Generative Algorithms: Challenges, Regulation, and the Future of Journalism
The technological revolution unfolding in newsrooms around the world is reshaping the very foundations of journalism. It is undermining long-standing publisher business models and confronting regulators with questions that still lack clear answers. With the adoption of the AI Act, the European Union has made a concerted attempt to set global standards.
Artificial intelligence is no longer confined to research labs or major technology corporations. Earlier forms of AI — predating the rise of generative systems — had already been used in newsrooms for years, supporting both content production and distribution. However, it was the emergence of widely accessible generative tools such as ChatGPT and Gemini that redirected public debate toward questions striking at the core of journalism: what constitutes truth, who can be considered the author, and whether information can be trusted when its sources remain opaque.
The data is unequivocal: as many as 67 percent of media organizations across 53 countries report using AI to some extent for content curation or article recommendation, according to the Reuters Institute for the Study of Journalism. This trend is reinforced by the Institute’s annual reports, which document the steady expansion of AI applications in both back-end automation and audience-facing services.
Generative AI: both promise and peril
Research into AI adoption in newsrooms reveals a clear disconnect between the scale of its use and journalists’ own level of understanding. A report by the Associated Press and the AI, Media & Democracy lab — based on a survey of experienced industry professionals — highlights the central importance of accountability as a core newsroom standard. It also points to strong interest in using AI to handle tedious and repetitive tasks, particularly those that can be reliably verified.
Advocates of emerging technologies emphasize the wide range of benefits generative AI can offer news organizations: automated summaries and newsletters, coverage of local events, topic discovery, and moderation of comment sections. These are labor-intensive, repetitive tasks — and automating them could free journalists to focus on work that truly requires reporting expertise and editorial judgment.
Yet the same technology brings significant risks. Generative AI systems not only make mistakes but can also produce highly convincing content and construct authoritative-sounding arguments in support of falsehoods. Existing licensing agreements between publishers and technology companies have largely failed to address this challenge.
AI can undoubtedly streamline “low-value” tasks and allow reporters to focus on more substantive work. However, critics warn that economic pressures may drive publishers to over-rely on automated content to increase output and reduce costs — potentially eroding both quality and originality.
Distinguishing between lawful and unlawful content is also becoming increasingly difficult. Tools designed to detect AI-generated material have so far proven largely ineffective. At the same time, the low barrier to entry has made AI an attractive tool for spammers mass-producing fake “news,” as well as for anti-democratic actors capable of launching scalable and potentially effective propaganda and disinformation campaigns.
Public trust: a complex puzzle
Public understanding of AI remains fragmented and often shaped by misconceptions. Research shows that news consumers frequently struggle to distinguish between AI-generated and human-created content. Moreover, some audiences perceive certain types of AI-generated news as less biased than human-written reporting — despite ample evidence that AI systems can reinforce societal biases and facilitate the spread of disinformation.
A cross-national study by the Reuters Institute, covering 28 countries and incorporating in-depth qualitative research from Mexico, the United Kingdom, and the United States, finds that audiences broadly accept the use of AI as a background tool — provided it remains under human oversight. Fully AI-generated content, however, is generally not accepted, while attitudes toward tasks such as reformatting or summarizing content fall somewhere in between.
In its report, CNTI issues an important warning to newsrooms: the public does not fully understand journalistic practices and needs this context to properly assess AI-related issues. As a result, transparency efforts must go beyond initial assumptions and include clearer communication about the role of human journalists.
The issue is further complicated by ongoing debates surrounding transparency requirements for algorithms used by technology companies.
Regulation: the EU AI Act as a global benchmark
Against a fragmented global landscape of regulatory approaches, the European Union stands out for its ambition to comprehensively govern artificial intelligence. Diverging national approaches reflect a broader lack of consensus on the values that should underpin AI legislation and ethical frameworks — a divide that complicates efforts to establish global regulatory standards.
Approaches vary widely. Some governments, such as those of Russia and China, directly or indirectly facilitate — and often control — AI development within their borders, granting them extensive access to citizens’ data. Others seek to foster innovation through multi-stakeholder engagement. The European Union, for its part, is actively pursuing a regulatory framework aimed at both enabling innovation and protecting citizens from potential harms.
Importantly, the influence of European legislation extends beyond its borders. The EU AI Act has already inspired legislative initiatives in countries such as Canada and Brazil.
Original article: https://focuseurope.pl/tematy/nowe-technologie/wiadomosci/wyzwania-regulacje-i-przyszlosc-mediow-w-erze-generatywnych-algorytmow