AI and quantum threats top security agendas

According to Thales, AI and Quantum threats have passed to the top of the anxiety list of wrestling institutions with data security. This is the main ready -made meals of the data threat report for the 2025 cyber security giant, the annual deep diving in the latest data security threats, emerging trends, and hot topics.

The results of this year are blatant: it is nearly seven organizations out of every ten organizations that now see the absolute speed of the development of artificial intelligence – especially when artificial intelligence is related to the first security headache related to its adoption. This anxiety is not only about speed; It is also fed due to the primary lack of integrity in artificial intelligence systems (reported by 64 % of those surveyed) and a disturbing deficit in trust (57 % anxiety).

AI Tolidi is a thirsty monster, and it depends greatly on high -quality information, often sensitive, for basic functions such as training models, making inferences, and of course the generation of content.

Since we offer rapid progress in “Agency AI” – systems that can act more independently – the pressure to ensure that high -caliber data quality is more important. After all, proper decisions and reliable actions of fully intelligent intelligence systems depend on the data that feeds on it.

Many organizations are already dived, as a third of the participants who refer to the Trucific intelligence indicate that they are either combined actively or are already a force to shift within their operations.

Security threats increase with organizations from artificial intelligence

Since obstetric intelligence casts a complex network of data safety challenges while providing strategic methods simultaneously to support defenses, its increasing integration indicates a distinct shift. Companies exceed merely dipping their toes in artificial intelligence water; They are now looking at the most mature operational publication.

Interestingly, although most of the respondents have exchanged Genai’s rapid absorption as their greatest security attention, those who tend to the artificial intelligence adoption curve do not strike the temporary stop button to completely close their systems or adjust their technology chimneys before moving forward. This rush of rapid shift – which often overwhelms efforts to ensure organizational readiness – may mean that these companies, perhaps inadvertently, create the most dangerous security places.

“The advanced rapid scene in GENAI presses institutions to move quickly, sometimes at the expense of caution, as they are racing to stay before the adoption curve,” said Erik Hanman, Senior analyst at the S& P Global Market Intelligence 451 Research.

“Many institutions publish Genai faster than they can fully understand their applications, which are doubled by the rapid spread of Saas tools that include Genai’s capabilities, adding layers of complexity and risk.”

In a more positive note, 73 % of the respondents are that they put money in Amnesty International’s security tools to counter threats, either through new budgets or by reformulating the current resources. Those who make artificial intelligence security a priority varied their methods: more than two -thirds have obtained tools from cloud service providers, and three out of five turns into firm security sellers, and half of nearly young people look forward to new or startups startups.

What he says in particular is the extent of the fast of the Turedi artificial intelligence to spending plans, which led to the second position in voting on the option in the rank, only amazed to this position through permanent interest in cloud safety. This shift strongly emphasizes the increasing recognition of the risks driven by artificial intelligence and the urgent need for specialized defenses to confront them.

Data violations show a modest decrease, although threats are still high

While the data breach is still waving on the horizon for many, its frequency has already decreased slightly over the past few years.

In 2021, 56 % of the institutions included in the survey said they were subjected to violation at some point; This number decreased to 45 % in the 2025 report. In the deepest, the percentage of the respondents who reported a breach over the past 12 months of 23 % in 2021 to 14 % fell more encouraged in 2025.

When it comes to the continuous wicked from the threat scene, harmful programs continue to lead the package, sticking to first place since 2021. The craftsman has risen to second place, prompting ransom programs to third place.

As for those who cause the most anxiety, external actors dominate this: the infiltrators are currently seen as the main danger, followed by the actors in the nation -state. A human error, while it is still an important factor, slipped to third place, below one position of the previous year.

Click on the sellers ready for quantum threats

The Thales 2025 data threat report also sheds light on disclosure of increased discomfort within most institutions on quantum security risks.

The highest threat here, which is martyred by 63 % of the respondents, is the danger that is waving on the horizon in a “compromise solution in the future”. This is an annoying possibility that strong quantum computers can shatter one day or even future encryption algorithms, which displays the data that has already been secured safely.

Hororate in its wake, select 61 % of the weaknesses in the main distribution, where quantum inventions can undermine the methods we use to share the encryption keys safely. Moreover, the 58 % “Harvest Now, Decrypt later” (HNDL) – a chilling scenario where the bodies can be deciphered where the encrypted data, which was obtained today, can be decoded today, by strong quantum machines in the future.

In response to these clouds, half of the survey organizations take a closer look at current encryption strategies with 60 % of the initial models already or evaluate post -quarter encryption solutions (PQC). However, confidence appears to be a rare commodity, as only a third hangs their hopes on the communications or clouds to move in this complex transition.

Todd Moore, Vice President of Data Security Products in Thales, commented: “The watch is moving on a post -quarter willingness. It is encouraging that three organizations out of every five organizations are already developing new zeros, but the publishing schedules may be narrow and leave important exposed data.

“Even with clear time schedules to move to PQC algorithms, the pace of change was slower than expected due to a mixture of old systems, complexity, and the challenge of balancing innovation with security.”

It is clear that there is a lot of work that must be done to obtain safety for truly operating data, not only to support the advanced capabilities of emerging technologies such as the Toulbia AI, but also to lay a safe basis for any brackets or lower threats.

(A picture of the Linforel house)

See also: Artificial intelligence tool accelerates government reactions, and experts urge caution

Do you want to learn more about cybersecurity and cloud from industry leaders? Chear Security & Cloud Expo, which is held in Amsterdam, California, and London. The comprehensive event was identified with other leading events including the Digital Transformation Week, the IOT Tech, Blockchain Expo and Ai & Big Data Expo.

Explore the next web events and seminars with which Techforge works here.